nips nips2011 nips2011-193 knowledge-graph by maker-knowledge-mining

193 nips-2011-Object Detection with Grammar Models

Source: pdf

Author: Ross B. Girshick, Pedro F. Felzenszwalb, David A. McAllester

Abstract: Compositional models provide an elegant formalism for representing the visual appearance of highly variable objects. While such models are appealing from a theoretical point of view, it has been difﬁcult to demonstrate that they lead to performance advantages on challenging datasets. Here we develop a grammar model for person detection and show that it outperforms previous high-performance systems on the PASCAL benchmark. Our model represents people using a hierarchy of deformable parts, variable structure and an explicit model of occlusion for partially visible objects. To train the model, we introduce a new discriminative framework for learning structured prediction models from weakly-labeled data. 1

Reference: text

Summary: the most important sentenses genereted by tfidf model

sentIndex sentText sentNum sentScore

1 Here we develop a grammar model for person detection and show that it outperforms previous high-performance systems on the PASCAL benchmark. [sent-11, score-0.749]

2 Our model represents people using a hierarchy of deformable parts, variable structure and an explicit model of occlusion for partially visible objects. [sent-12, score-0.535]

3 However, it has been difﬁcult to demonstrate that sophisticated compositional models lead to performance advantages on challenging metrics such as the PASCAL object detection benchmark [9]. [sent-16, score-0.288]

4 In this paper we achieve new levels of performance for person detection using a grammar model that is richer than previous models used in high-performance systems. [sent-17, score-0.749]

5 Our models are based on the object detection grammar formalism in [11]. [sent-19, score-0.755]

6 Deformation rules allow for the parts of an object to move relative to each other, leading to hierarchical deformable part models. [sent-21, score-0.298]

7 Structural variability provides choice between multiple part subtypes — effectively creating mixture models throughout the compositional hierarchy — and also enables optional parts. [sent-22, score-0.306]

8 In this formalism parts may be reused both within an object category and across object categories. [sent-23, score-0.314]

9 For example, each component has a head part trained independently from the head part of the other components. [sent-28, score-0.254]

10 Here we construct a single grammar model that allows more ﬂexibility in describing the amount of the person that is visible. [sent-29, score-0.627]

11 The grammar model avoids dividing the training data between different components and thus uses the training data more efﬁciently. [sent-30, score-0.622]

12 The parts in the model, such as the head part, are shared across different interpretations of the degree of visibility of the person. [sent-31, score-0.261]

13 The grammar model also includes subtype choice at the part level to accommodate greater appearance 1 variability across object instances. [sent-32, score-0.811]

14 We use parts with subparts to beneﬁt from high-resolution image data, while also allowing for deformations. [sent-33, score-0.333]

15 Unlike previous approaches, we explicitly model the source of occlusion for partially visible objects. [sent-34, score-0.408]

16 However, our training method, image features, and grammar design are substantially different. [sent-39, score-0.596]

17 There has been recent related work on occlusion modeling in pedestrian and person images [7, 18]. [sent-41, score-0.512]

18 assume access to depth and motion information in order to estimate occlusion boundaries. [sent-43, score-0.336]

19 rely on the observation that the scores of individual ﬁlter cells (using the Dalal and Triggs detector [5]) can reliably predict occlusion in the INRIA pedestrian data. [sent-45, score-0.474]

20 In addition to developing a grammar model for detecting people, we develop new training methods which contribute to our boost in performance. [sent-47, score-0.612]

21 Training data for vision is often assigned weak labels such as bounding boxes or just the names of objects occurring in the image. [sent-48, score-0.219]

22 Existing structured prediction methods, such as structural SVM [16, 17] and latent structural SVM [19], do not directly support weak labels together with strong predictions. [sent-50, score-0.292]

23 We develop the notion of a weaklabel structural SVM which generalizes structural SVMs and latent-structural SVMs. [sent-51, score-0.188]

24 2 Grammar models Object detection grammars [11] represent objects recursively in terms of other objects. [sent-56, score-0.19]

25 The structure of a grammar model is deﬁned by a set, R, of weighted productions of the form s X(ω0 ) −→ { Y1 (ω1 ), . [sent-62, score-0.588]

26 We can expand a placed nonterminal to a bag of placed terminals by repeatedly applying productions. [sent-67, score-0.252]

27 The leaves of T are labeled with placed terminals, and the internal nodes of T are labeled with placed nonterminals and with the productions used to replace those symbols. [sent-69, score-0.236]

28 We deﬁne appearance models for the terminals using a function score(A, ω) that computes a score for placing the terminal A at location ω. [sent-70, score-0.356]

29 We deﬁne the score of a derivation tree T to be the sum of the scores of the productions used to generate T , plus the score of placing the terminals associated with T ’s leaves in their respective locations. [sent-72, score-0.484]

30 We deﬁne the appearance models for terminals by associating a ﬁlter FA with each terminal A. [sent-74, score-0.235]

31 2 Subtype 1 Subtype 2 Example detections and derived ﬁlters Part 1 Part 2 Part 3 Part 4 Part 5 Part 6 Occluder Parts 1-6 (no occlusion) Parts 1-4 & occluder Parts 1-2 & occluder Figure 1: Shallow grammar model. [sent-77, score-0.708]

32 This ﬁgure illustrates a shallow version of our grammar model (Section 2. [sent-78, score-0.575]

33 This model has six person parts and an occlusion model (“occluder”), each of which comes in one of two subtypes. [sent-80, score-0.583]

34 A detection places one subtype of each visible part at a location and scale in the image. [sent-81, score-0.385]

35 We consider models with productions speciﬁed by two kinds of schemas (a schema is a template for generating productions). [sent-86, score-0.259]

36 A structure schema speciﬁes one production for each placement ω ∈ Ω, s X(ω) −→ { Y1 (ω ⊕ δ1 ), . [sent-87, score-0.2]

37 A deformation schema speciﬁes one production for each placement ω ∈ Ω and displacement δ ∈ ∆, α·φ(δ) X(ω) −→ { Y (ω ⊕ δ) }. [sent-94, score-0.346]

38 The parameters of our models are deﬁned by a weight vector w with entries for the score of each structure schema, the deformation parameters of each deformation schema and the ﬁlter coefﬁcients associated with each terminal. [sent-98, score-0.412]

39 Then score(T ) = w · Φ(T ), where Φ(T ) is the sum of (sparse) feature vectors associated with each placed terminal and production in T . [sent-99, score-0.191]

40 1 A grammar model for detecting people Each component in the person model learned by the voc-release4 system [12] is tuned to detect people under a prototypical visibility pattern. [sent-101, score-0.924]

41 Based on this observation we designed, by hand, the structure of a grammar that models visibility by using structural variability and optional parts. [sent-102, score-0.732]

42 After explaining this model, we describe a deeper model that includes deformable subparts at higher resolutions. [sent-104, score-0.283]

43 Fine-grained occlusion Our grammar model has a start symbol Q that can be expanded using one of six possible structure schemas. [sent-105, score-0.917]

44 These choices model different degrees of visibility ranging from heavy occlusion (only the head and shoulders are visible) to no occlusion at all. [sent-106, score-0.912]

45 Beyond modeling ﬁne-grained occlusion patterns when compared to the mixture models from [7] and [12], our grammar model is also richer in a number of ways. [sent-107, score-0.867]

46 3 Occlusion model If a person is occluded, then there must be some cause of the occlusion — either the edge of the image or an occluding object, such as a desk or dinner table. [sent-109, score-0.51]

47 Part subtypes The mixture model from [12] has two subtypes for each mixture component. [sent-111, score-0.298]

48 Our grammar model has two subtypes for each part, which are also forced to be mirror images of each other. [sent-113, score-0.636]

49 But in the case of our grammar model, the decision of which part subtype to instantiate at detection time is independent for each part. [sent-114, score-0.777]

50 The shallow person grammar model is deﬁned by the following grammar. [sent-115, score-0.71]

51 , Y6 (ω ⊕ δ6 ) } 0 Yp (ω) −→ { Yp,t (ω) } Yp,t (ω) αp,t ·φ(δ) −→ { Ap,t (ω ⊕ δ) } αt ·φ(δ) 0 O(ω) −→ { Ot (ω) } Ot (ω) −→ { At (ω ⊕ δ) } The grammar has a start symbol Q with six alternate choices that derive people under varying degrees of visibility (occlusion). [sent-129, score-0.761]

52 A derivation selects a subtype and displacement for each visible part. [sent-132, score-0.25]

53 The parameters of the grammar (production scores, deformation parameters and ﬁlters) are learned with the discriminative procedure described in Section 4. [sent-133, score-0.61]

54 Deeper model We extend the shallow model by adding deformable subparts at two scales: (1) the same as, and (2) twice the resolution of the start symbol Q. [sent-135, score-0.419]

55 When detecting large objects, high-resolution subparts capture ﬁne image details. [sent-136, score-0.312]

56 However, when detecting small objects, highresolution subparts cannot be used because they “fall off the bottom” of the feature map pyramid. [sent-137, score-0.314]

57 The model uses derivations with low-resolution subparts when detecting small objects. [sent-138, score-0.322]

58 We begin by replacing the productions from Yp,t in the grammar above, and then adding new productions. [sent-139, score-0.588]

59 , Np }, where Np is the number of subparts in a top-level part Yp . [sent-144, score-0.278]

60 The part terminal Ap,t can move relative to Q and the subpart terminal Ap,t,r,u can move relative to Ap,t . [sent-149, score-0.23]

61 The displacements δp,t,H,u place the symbols Wp,t,H,u one octave below Zp,t in the feature map pyramid. [sent-150, score-0.208]

62 We add subparts to the ﬁrst two top-level parts (p = 1 and 2), with the number of subparts set to N1 = 3 and N2 = 2. [sent-152, score-0.512]

63 We ﬁnd that adding additional subparts does not improve detection performance. [sent-153, score-0.34]

64 We deﬁne box(T ) by assigning a detection window size, in feature map coordinates, to each structure schema that can be applied to Q. [sent-161, score-0.246]

65 The absolute location and size of a detection depends on the placement of Q. [sent-163, score-0.195]

66 For the ﬁrst ﬁve production schemas, the ideal location of the occlusion part, O, is outside of box(T ). [sent-164, score-0.436]

67 A simple example of a weakly-labeled training problem comes from learning sliding window classiﬁers in the PASCAL object detection dataset. [sent-179, score-0.284]

68 The training data speciﬁes pixel-accurate bounding boxes for the target objects while a sliding window classiﬁer reports boxes with a ﬁxed aspect ratio and at a ﬁnite number of scales. [sent-180, score-0.212]

69 We deﬁne a weak-label structural SVM (WL-SSVM) by the following training equation, E(w) = 1 ||w||2 + C 2 n L (w, xi , yi ). [sent-184, score-0.222]

70 But even in this case a ˆ ˆ structural SVM pushes the score w · Φ(x, y) to be above w · Φ(x, y ). [sent-200, score-0.215]

71 4 Training grammar models Now we consider learning the parameters of an object detection grammar using the training data in the PASCAL VOC datasets with the WL-SSVM framework. [sent-207, score-1.268]

72 The training data speciﬁes a bounding box for each instance of an object in a set of training images. [sent-213, score-0.393]

73 For each training image I, and for each bounding box B in I, we deﬁne a foreground example (x, y), where y = B, x speciﬁes the image I, and the set of valid predictions S(x) includes: 1. [sent-218, score-0.372]

74 The overlap requirements in (1) ensure that we consider only predictions that are relevant for a particular object instance, while avoiding interactions with other objects in the image. [sent-224, score-0.223]

75 1 Loss functions The PASCAL benchmark requires a correct detection to have at least 50% overlap with a groundtruth bounding box. [sent-231, score-0.259]

76 For a foreground example this pushes down the score of detections that don’t overlap with the bounding box label by at least 50%. [sent-236, score-0.478]

77 For a foreground example this ensures that the maximizer of (6b) is a detection with high overlap with the bounding box label. [sent-239, score-0.409]

78 2 After computing si (wt ) and Φ(xi , si (wt )) for all examples (implicitly for background examples), the weight vector is updated by minimizing a convex upper bound on the objective E(w): wt+1 = 1 argmin ||w||2 + C 2 w n max [w · Φ(xi , s) + Lmargin (yi , s)] − w · Φ(xi , si (wt )) . [sent-268, score-0.253]

79 Instead of applying data mining to the foreground examples, each time we compute si (wt ) for a foreground example, we also compute the top M scoring outputs s ∈ S(xi ) of wt · Φ(xi , s) + Lmargin (yi , s), and place the corresponding feature vectors in the data mining cache. [sent-274, score-0.39]

80 After training the initial ﬁlter, we slice it into subﬁlters (one 8 × 8 and ﬁve 3 × 8) that form the building blocks of the grammar model. [sent-286, score-0.557]

81 We mirror these six ﬁlters to get subtypes, and then add subparts using the energy covering heuristic in [10, 12]. [sent-287, score-0.288]

82 5 Experimental results We evaluated the performance of our person grammar and training framework on the PASCAL VOC 2007 and 2010 datasets [8, 9]. [sent-288, score-0.692]

83 We report the AP score of the UoC-TTI “raw” person detector, as well as the scores after applying the bounding 7 (a) Full visibility (b) Occlusion boundaries (c) Early termination (d) Mistakes Figure 2: Example detections. [sent-294, score-0.471]

84 (b) Examples where the occlusion part detects an occlusion boundary. [sent-298, score-0.732]

85 (d) Mistakes where the model did not detect occlusion properly. [sent-300, score-0.336]

86 Comparing raw detector outputs our grammar model signiﬁcantly outperforms the mixture model: 47. [sent-302, score-0.632]

87 We also applied the two post-processing steps to the grammar model, and found that unlike with the mixture model, the grammar model does not beneﬁt from bounding box prediction. [sent-306, score-1.189]

88 This is likely because our ﬁne-grained occlusion model reduces the number of near misses that are ﬁxed by bounding box prediction. [sent-307, score-0.502]

89 To test context rescoring, we used the UoC-TTI detection data for the other 19 object classes. [sent-308, score-0.219]

90 Their system requires detailed pose and visibility annotations, in contrast to our grammar model which was trained only with bounding box labels. [sent-312, score-0.776]

91 Prior to context rescoring, our model scores one point lower than the poselets model, and after rescoring it scores one point higher. [sent-313, score-0.242]

92 WL-SSVM improves performance of the grammar model by 1. [sent-318, score-0.492]

93 To investigate model structure, we evaluated the effect of part subtypes and occlusion modeling. [sent-322, score-0.506]

94 Removing subtypes reduces the score of the grammar model from 46. [sent-323, score-0.695]

95 Removing the occlusion part also decreases the score from 46. [sent-326, score-0.489]

96 6 Discussion Our results establish grammar-based methods as a high-performance approach to object detection by demonstrating their effectiveness on the challenging task of detecting people in the PASCAL VOC datasets. [sent-331, score-0.336]

97 To do this, we carefully designed a ﬂexible grammar model that can detect people under a wide range of partial occlusion, pose, and appearance variability. [sent-332, score-0.613]

98 Automatically learning the structure of grammar models remains a signiﬁcant challenge for future work. [sent-333, score-0.492]

99 We hope that our empirical success will provide motivation for pursing this goal, and that the structure of our handcrafted grammar will yield insights into the properties that an automatically learned grammar might require. [sent-334, score-0.984]

100 We also develop a structured training framework, weak-label structural SVM, that naturally handles learning a model with strong outputs, such as derivation trees, from data with weak labels, such as bounding boxes. [sent-335, score-0.324]

similar papers computed by tfidf model

tfidf for this paper:

wordName wordTfidf (topN-words)

[('grammar', 0.492), ('occlusion', 0.336), ('loutput', 0.309), ('subparts', 0.218), ('lmargin', 0.2), ('pascal', 0.139), ('person', 0.135), ('detection', 0.122), ('deformation', 0.118), ('visibility', 0.118), ('subtypes', 0.11), ('terminals', 0.109), ('subtype', 0.103), ('object', 0.097), ('productions', 0.096), ('structural', 0.094), ('score', 0.093), ('rescoring', 0.091), ('lsvm', 0.088), ('box', 0.087), ('shallow', 0.083), ('schema', 0.083), ('schemas', 0.08), ('bounding', 0.079), ('background', 0.079), ('voc', 0.076), ('parts', 0.076), ('occluder', 0.073), ('production', 0.072), ('visible', 0.072), ('detections', 0.07), ('wt', 0.069), ('compositional', 0.069), ('displacements', 0.069), ('objects', 0.068), ('head', 0.067), ('terminal', 0.067), ('deformable', 0.065), ('training', 0.065), ('foreground', 0.063), ('people', 0.062), ('part', 0.06), ('svm', 0.06), ('appearance', 0.059), ('poselets', 0.059), ('overlap', 0.058), ('si', 0.058), ('detecting', 0.055), ('econcave', 0.055), ('shoulders', 0.055), ('symbol', 0.053), ('placed', 0.052), ('detector', 0.051), ('cccp', 0.05), ('outputs', 0.05), ('derivations', 0.049), ('derivation', 0.047), ('scores', 0.046), ('ap', 0.045), ('placement', 0.045), ('formalism', 0.044), ('lters', 0.044), ('chicago', 0.044), ('zhu', 0.043), ('yp', 0.041), ('ramp', 0.041), ('pedestrian', 0.041), ('map', 0.041), ('weak', 0.039), ('nonterminal', 0.039), ('image', 0.039), ('mixture', 0.039), ('loss', 0.038), ('mcallester', 0.038), ('felzenszwalb', 0.037), ('six', 0.036), ('argmaxs', 0.036), ('bbox', 0.036), ('econvex', 0.036), ('enzweiler', 0.036), ('nonterminals', 0.036), ('octave', 0.036), ('subpart', 0.036), ('lter', 0.035), ('mirror', 0.034), ('labels', 0.033), ('girshick', 0.033), ('symbols', 0.033), ('xi', 0.032), ('latent', 0.032), ('yi', 0.031), ('hog', 0.031), ('ranking', 0.03), ('softer', 0.029), ('place', 0.029), ('mining', 0.029), ('location', 0.028), ('optional', 0.028), ('displacement', 0.028), ('pushes', 0.028)]

similar papers list:

simIndex simValue paperId paperTitle

same-paper 1 1.0000002 193 nips-2011-Object Detection with Grammar Models

Author: Ross B. Girshick, Pedro F. Felzenszwalb, David A. McAllester

2 0.23753421 127 nips-2011-Image Parsing with Stochastic Scene Grammar

Author: Yibiao Zhao, Song-chun Zhu

Abstract: This paper proposes a parsing algorithm for scene understanding which includes four aspects: computing 3D scene layout, detecting 3D objects (e.g. furniture), detecting 2D faces (windows, doors etc.), and segmenting background. In contrast to previous scene labeling work that applied discriminative classiﬁers to pixels (or super-pixels), we use a generative Stochastic Scene Grammar (SSG). This grammar represents the compositional structures of visual entities from scene categories, 3D foreground/background, 2D faces, to 1D lines. The grammar includes three types of production rules and two types of contextual relations. Production rules: (i) AND rules represent the decomposition of an entity into sub-parts; (ii) OR rules represent the switching among sub-types of an entity; (iii) SET rules represent an ensemble of visual entities. Contextual relations: (i) Cooperative “+” relations represent positive links between binding entities, such as hinged faces of a object or aligned boxes; (ii) Competitive “-” relations represents negative links between competing entities, such as mutually exclusive boxes. We design an efﬁcient MCMC inference algorithm, namely Hierarchical cluster sampling, to search in the large solution space of scene conﬁgurations. The algorithm has two stages: (i) Clustering: It forms all possible higher-level structures (clusters) from lower-level entities by production rules and contextual relations. (ii) Sampling: It jumps between alternative structures (clusters) in each layer of the hierarchy to ﬁnd the most probable conﬁguration (represented by a parse tree). In our experiment, we demonstrate the superiority of our algorithm over existing methods on public dataset. In addition, our approach achieves richer structures in the parse tree. 1

3 0.18409877 154 nips-2011-Learning person-object interactions for action recognition in still images

Author: Vincent Delaitre, Josef Sivic, Ivan Laptev

Abstract: We investigate a discriminatively trained model of person-object interactions for recognizing common human actions in still images. We build on the locally order-less spatial pyramid bag-of-features model, which was shown to perform extremely well on a range of object, scene and human action recognition tasks. We introduce three principal contributions. First, we replace the standard quantized local HOG/SIFT features with stronger discriminatively trained body part and object detectors. Second, we introduce new person-object interaction features based on spatial co-occurrences of individual body parts and objects. Third, we address the combinatorial problem of a large number of possible interaction pairs and propose a discriminative selection procedure using a linear support vector machine (SVM) with a sparsity inducing regularizer. Learning of action-speciﬁc body part and object interactions bypasses the difﬁcult problem of estimating the complete human body pose conﬁguration. Beneﬁts of the proposed model are shown on human action recognition in consumer photographs, outperforming the strong bag-of-features baseline. 1

4 0.1571286 180 nips-2011-Multiple Instance Filtering

Author: Kamil A. Wnuk, Stefano Soatto

Abstract: We propose a robust ﬁltering approach based on semi-supervised and multiple instance learning (MIL). We assume that the posterior density would be unimodal if not for the eﬀect of outliers that we do not wish to explicitly model. Therefore, we seek for a point estimate at the outset, rather than a generic approximation of the entire posterior. Our approach can be thought of as a combination of standard ﬁnite-dimensional ﬁltering (Extended Kalman Filter, or Unscented Filter) with multiple instance learning, whereby the initial condition comes with a putative set of inlier measurements. We show how both the state (regression) and the inlier set (classiﬁcation) can be estimated iteratively and causally by processing only the current measurement. We illustrate our approach on visual tracking problems whereby the object of interest (target) moves and evolves as a result of occlusions and deformations, and partial knowledge of the target is given in the form of a bounding box (training set). 1

5 0.12742704 166 nips-2011-Maximal Cliques that Satisfy Hard Constraints with Application to Deformable Object Model Learning

Author: Xinggang Wang, Xiang Bai, Xingwei Yang, Wenyu Liu, Longin J. Latecki

Abstract: We propose a novel inference framework for ﬁnding maximal cliques in a weighted graph that satisfy hard constraints. The constraints specify the graph nodes that must belong to the solution as well as mutual exclusions of graph nodes, i.e., sets of nodes that cannot belong to the same solution. The proposed inference is based on a novel particle ﬁlter algorithm with state permeations. We apply the inference framework to a challenging problem of learning part-based, deformable object models. Two core problems in the learning framework, matching of image patches and ﬁnding salient parts, are formulated as two instances of the problem of ﬁnding maximal cliques with hard constraints. Our learning framework yields discriminative part based object models that achieve very good detection rate, and outperform other methods on object classes with large deformation. 1

6 0.12654389 1 nips-2011-$\theta$-MRF: Capturing Spatial and Semantic Structure in the Parameters for Scene Understanding

7 0.12051287 290 nips-2011-Transfer Learning by Borrowing Examples for Multiclass Object Detection

8 0.11179025 233 nips-2011-Rapid Deformable Object Detection using Dual-Tree Branch-and-Bound

9 0.10195859 303 nips-2011-Video Annotation and Tracking with Active Learning

10 0.09205737 126 nips-2011-Im2Text: Describing Images Using 1 Million Captioned Photographs

11 0.088027731 247 nips-2011-Semantic Labeling of 3D Point Clouds for Indoor Scenes

12 0.08322081 168 nips-2011-Maximum Margin Multi-Instance Learning

13 0.079503424 138 nips-2011-Joint 3D Estimation of Objects and Scene Layout

14 0.078940012 275 nips-2011-Structured Learning for Cell Tracking

15 0.07716886 223 nips-2011-Probabilistic Joint Image Segmentation and Labeling

16 0.075209133 169 nips-2011-Maximum Margin Multi-Label Structured Prediction

17 0.071615502 103 nips-2011-Generalization Bounds and Consistency for Latent Structural Probit and Ramp Loss

18 0.064628989 231 nips-2011-Randomized Algorithms for Comparison-based Search

19 0.063704982 91 nips-2011-Exploiting spatial overlap to efficiently compute appearance distances between image windows

20 0.061741006 141 nips-2011-Large-Scale Category Structure Aware Image Categorization

similar papers computed by lsi model

lsi for this paper:

topicId topicWeight

[(0, 0.179), (1, 0.099), (2, -0.116), (3, 0.168), (4, 0.103), (5, 0.042), (6, -0.002), (7, -0.07), (8, -0.011), (9, 0.092), (10, 0.083), (11, -0.078), (12, -0.035), (13, 0.026), (14, 0.004), (15, -0.011), (16, 0.026), (17, 0.018), (18, -0.041), (19, 0.065), (20, -0.026), (21, -0.042), (22, -0.061), (23, 0.11), (24, 0.063), (25, -0.055), (26, -0.067), (27, 0.113), (28, -0.027), (29, 0.011), (30, -0.045), (31, 0.064), (32, 0.045), (33, 0.094), (34, 0.019), (35, 0.057), (36, -0.01), (37, 0.013), (38, -0.009), (39, -0.034), (40, -0.038), (41, -0.164), (42, 0.099), (43, -0.045), (44, -0.031), (45, 0.089), (46, -0.011), (47, 0.052), (48, -0.05), (49, 0.025)]

similar papers list:

simIndex simValue paperId paperTitle

same-paper 1 0.92561996 193 nips-2011-Object Detection with Grammar Models

Author: Ross B. Girshick, Pedro F. Felzenszwalb, David A. McAllester

2 0.71835351 290 nips-2011-Transfer Learning by Borrowing Examples for Multiclass Object Detection

Author: Joseph J. Lim, Antonio Torralba, Ruslan Salakhutdinov

Abstract: Despite the recent trend of increasingly large datasets for object detection, there still exist many classes with few training examples. To overcome this lack of training data for certain classes, we propose a novel way of augmenting the training data for each class by borrowing and transforming examples from other classes. Our model learns which training instances from other classes to borrow and how to transform the borrowed examples so that they become more similar to instances from the target class. Our experimental results demonstrate that our new object detector, with borrowed and transformed examples, improves upon the current state-of-the-art detector on the challenging SUN09 object detection dataset. 1

3 0.71716052 127 nips-2011-Image Parsing with Stochastic Scene Grammar

Author: Yibiao Zhao, Song-chun Zhu

4 0.7152316 154 nips-2011-Learning person-object interactions for action recognition in still images

Author: Vincent Delaitre, Josef Sivic, Ivan Laptev

5 0.70306736 138 nips-2011-Joint 3D Estimation of Objects and Scene Layout

Author: Andreas Geiger, Christian Wojek, Raquel Urtasun

Abstract: We propose a novel generative model that is able to reason jointly about the 3D scene layout as well as the 3D location and orientation of objects in the scene. In particular, we infer the scene topology, geometry as well as trafﬁc activities from a short video sequence acquired with a single camera mounted on a moving car. Our generative model takes advantage of dynamic information in the form of vehicle tracklets as well as static information coming from semantic labels and geometry (i.e., vanishing points). Experiments show that our approach outperforms a discriminative baseline based on multiple kernel learning (MKL) which has access to the same image information. Furthermore, as we reason about objects in 3D, we are able to signiﬁcantly increase the performance of state-of-the-art object detectors in their ability to estimate object orientation. 1

6 0.668859 180 nips-2011-Multiple Instance Filtering

7 0.64228648 247 nips-2011-Semantic Labeling of 3D Point Clouds for Indoor Scenes

8 0.60933369 166 nips-2011-Maximal Cliques that Satisfy Hard Constraints with Application to Deformable Object Model Learning

9 0.6088109 1 nips-2011-$\theta$-MRF: Capturing Spatial and Semantic Structure in the Parameters for Scene Understanding

10 0.56618655 35 nips-2011-An ideal observer model for identifying the reference frame of objects

11 0.56054068 275 nips-2011-Structured Learning for Cell Tracking

12 0.51727307 233 nips-2011-Rapid Deformable Object Detection using Dual-Tree Branch-and-Bound

13 0.51669389 304 nips-2011-Why The Brain Separates Face Recognition From Object Recognition

14 0.4803488 303 nips-2011-Video Annotation and Tracking with Active Learning

15 0.4573139 126 nips-2011-Im2Text: Describing Images Using 1 Million Captioned Photographs

16 0.42084232 232 nips-2011-Ranking annotators for crowdsourced labeling tasks

17 0.41692626 91 nips-2011-Exploiting spatial overlap to efficiently compute appearance distances between image windows

18 0.40359369 169 nips-2011-Maximum Margin Multi-Label Structured Prediction

19 0.39506337 223 nips-2011-Probabilistic Joint Image Segmentation and Labeling

20 0.38518238 293 nips-2011-Understanding the Intrinsic Memorability of Images

similar papers computed by lda model

lda for this paper:

topicId topicWeight

[(0, 0.023), (4, 0.483), (20, 0.055), (26, 0.016), (31, 0.05), (33, 0.041), (43, 0.052), (45, 0.073), (57, 0.029), (74, 0.04), (83, 0.019), (99, 0.03)]

similar papers list:

simIndex simValue paperId paperTitle

1 0.93372732 130 nips-2011-Inductive reasoning about chimeric creatures

Author: Charles Kemp

Abstract: Given one feature of a novel animal, humans readily make inferences about other features of the animal. For example, winged creatures often ﬂy, and creatures that eat ﬁsh often live in the water. We explore the knowledge that supports these inferences and compare two approaches. The ﬁrst approach proposes that humans rely on abstract representations of dependency relationships between features, and is formalized here as a graphical model. The second approach proposes that humans rely on speciﬁc knowledge of previously encountered animals, and is formalized here as a family of exemplar models. We evaluate these models using a task where participants reason about chimeras, or animals with pairs of features that have not previously been observed to co-occur. The results support the hypothesis that humans rely on explicit representations of relationships between features. Suppose that an eighteenth-century naturalist learns about a new kind of animal that has fur and a duck’s bill. Even though the naturalist has never encountered an animal with this pair of features, he should be able to make predictions about other features of the animal—for example, the animal could well live in water but probably does not have feathers. Although the platypus exists in reality, from a eighteenth-century perspective it qualiﬁes as a chimera, or an animal that combines two or more features that have not previously been observed to co-occur. Here we describe a probabilistic account of inductive reasoning and use it to account for human inferences about chimeras. The inductive problems we consider are special cases of the more general problem in Figure 1a where a reasoner is given a partially observed matrix of animals by features then asked to infer the values of the missing entries. This general problem has been previously studied and is addressed by computational models of property induction, categorization, and generalization [1–7]. A challenge faced by all of these models is to capture the background knowledge that guides inductive inferences. Some accounts rely on similarity relationships between animals [6, 8], others rely on causal relationships between features [9, 10], and others incorporate relationships between animals and relationships between features [11]. We will evaluate graphical models that capture both kinds of relationships (Figure 1a), but will focus in particular on relationships between features. Psychologists have previously suggested that humans rely on explicit mental representations of relationships between features [12–16]. Often these representations are described as theories—for example, theories that specify a causal relationship between having wings and ﬂying, or living in the sea and eating ﬁsh. Relationships between features may take several forms: for example, one feature may cause, enable, prevent, be inconsistent with, or be a special case of another feature. For simplicity, we will treat all of these relationships as instances of dependency relationships between features, and will capture them using an undirected graphical model. Previous studies have used graphical models to account for human inferences about features but typically these studies consider toy problems involving a handful of novel features such as “has gene X14” or “has enzyme Y132” [9, 11]. Participants might be told, for example, that gene X14 leads to the production of enzyme Y132, then asked to use this information when reasoning about novel animals. Here we explore whether a graphical model approach can account for inferences 1 (a) slow heavy flies (b) wings hippo 1 1 0 0 rhino 1 1 0 0 sparrow 0 0 1 1 robin 0 0 1 1 new ? ? 1 ? o Figure 1: Inductive reasoning about animals and features. (a) Inferences about the features of a new animal onew that ﬂies may draw on similarity relationships between animals (the new animal is similar to sparrows and robins but not hippos and rhinos), and on dependency relationships between features (ﬂying and having wings are linked). (b) A graph product produced by combining the two graph structures in (a). about familiar features. Working with familiar features raises a methodological challenge since participants have a substantial amount of knowledge about these features and can reason about them in multiple ways. Suppose, for example, that you learn that a novel animal can ﬂy (Figure 1a). To conclude that the animal probably has wings, you might consult a mental representation similar to the graph at the top of Figure 1a that speciﬁes a dependency relationship between ﬂying and having wings. On the other hand, you might reach the same conclusion by thinking about ﬂying creatures that you have previously encountered (e.g. sparrows and robins) and noticing that these creatures have wings. Since the same conclusion can be reached in two different ways, judgments about arguments of this kind provide little evidence about the mental representations involved. The challenge of working with familiar features directly motivates our focus on chimeras. Inferences about chimeras draw on rich background knowledge but require the reasoner to go beyond past experience in a fundamental way. For example, if you learn that an animal ﬂies and has no legs, you cannot make predictions about the animal by thinking of ﬂying, no-legged creatures that you have previously encountered. You may, however, still be able to infer that the novel animal has wings if you understand the relationship between ﬂying and having wings. We propose that graphical models over features can help to explain how humans make inferences of this kind, and evaluate our approach by comparing it to a family of exemplar models. The next section introduces these models, and we then describe two experiments designed to distinguish between the models. 1 Reasoning about objects and features Our models make use of a binary matrix D where the rows {o1 , . . . , o129 } correspond to objects, and the columns {f 1 , . . . , f 56 } correspond to features. A subset of the objects is shown in Figure 2a, and the full set of features is shown in Figure 2b and its caption. Matrix D was extracted from the Leuven natural concept database [17], which includes 129 animals and 757 features in total. We chose a subset of these features that includes a mix of perceptual and behavioral features, and that includes many pairs of features that depend on each other. For example, animals that “live in water” typically “can swim,” and animals that have “no legs” cannot “jump far.” Matrix D can be used to formulate problems where a reasoner observes one or two features of a new object (i.e. animal o130 ) and must make inferences about the remaining features of the animal. The next two sections describe graphical models that can be used to address this problem. The ﬁrst graphical model O captures relationships between objects, and the second model F captures relationships between features. We then discuss how these models can be combined, and introduce a family of exemplar-style models that will be compared with our graphical models. A graphical model over objects Many accounts of inductive reasoning focus on similarity relationships between objects [6, 8]. Here we describe a tree-structured graphical model O that captures these relationships. The tree was constructed from matrix D using average linkage clustering and the Jaccard similarity measure, and part of the resulting structure is shown in Figure 2a. The subtree in Figure 2a includes clusters 2 alligator caiman crocodile monitor lizard dinosaur blindworm boa cobra python snake viper chameleon iguana gecko lizard salamander frog toad tortoise turtle anchovy herring sardine cod sole salmon trout carp pike stickleback eel flatfish ray plaice piranha sperm whale squid swordfish goldfish dolphin orca whale shark bat fox wolf beaver hedgehog hamster squirrel mouse rabbit bison elephant hippopotamus rhinoceros lion tiger polar bear deer dromedary llama giraffe zebra kangaroo monkey cat dog cow horse donkey pig sheep (a) (b) can swim lives in water eats fish eats nuts eats grain eats grass has gills can jump far has two legs has no legs has six legs has four legs can fly can be ridden has sharp teeth nocturnal has wings strong predator can see in dark eats berries lives in the sea lives in the desert crawls lives in the woods has mane lives in trees can climb well lives underground has feathers has scales slow has fur heavy Figure 2: Graph structures used to deﬁne graphical models O and F. (a) A tree that captures similarity relationships between animals. The full tree includes 129 animals, and only part of the tree is shown here. The grey points along the branches indicate locations where a novel animal o130 could be attached to the tree. (b) A network capturing pairwise dependency relationships between features. The edges capture both positive and negative dependencies. All edges in the network are shown, and the network also includes 20 isolated nodes for the following features: is black, is blue, is green, is grey, is pink, is red, is white, is yellow, is a pet, has a beak, stings, stinks, has a long neck, has feelers, sucks blood, lays eggs, makes a web, has a hump, has a trunk, and is cold-blooded. corresponding to amphibians and reptiles, aquatic creatures, and land mammals, and the subtree omitted for space includes clusters for insects and birds. We assume that the features in matrix D (i.e. the columns) are generated independently over O: P (f i |O, π i , λi ). P (D|O, π, λ) = i i i i The distribution P (f |O, π , λ ) is based on the intuition that nearby nodes in O tend to have the same value of f i . Previous researchers [8, 18] have used a directed graphical model where the distribution at the root node is based on the baserate π i , and any other node v with parent u has the following conditional probability distribution: i P (v = 1|u) = π i + (1 − π i )e−λ l , if u = 1 i π i − π i e−λ l , if u = 0 (1) where l is the length of the branch joining node u to node v. The variability parameter λi captures the extent to which feature f i is expected to vary over the tree. Note, for example, that any node v must take the same value as its parent u when λ = 0. To avoid free parameters, the feature baserates π i and variability parameters λi are set to their maximum likelihood values given the observed values of the features {f i } in the data matrix D. The conditional distributions in Equation 1 induce a joint distribution over all of the nodes in graph O, and the distribution P (f i |O, π i , λi ) is computed by marginalizing out the values of the internal nodes. Although we described O as a directed graphical model, the model can be converted into an equivalent undirected model with a potential for each edge in the tree and a potential for the root node. Here we use the undirected version of the model, which is a natural counterpart to the undirected model F described in the next section. The full version of structure O in Figure 2a includes 129 familiar animals, and our task requires inferences about a novel animal o130 that must be slotted into the structure. Let D′ be an expanded version of D that includes a row for o130 , and let O′ be an expanded version of O that includes a node for o130 . The edges in Figure 2a are marked with evenly spaced gray points, and we use a 3 uniform prior P (O′ ) over all trees that can be created by attaching o130 to one of these points. Some of these trees have identical topologies, since some edges in Figure 2a have multiple gray points. Predictions about o130 can be computed using: P (D′ |D) = P (D′ |O′ , D)P (O′ |D) ∝ O′ P (D′ |O′ , D)P (D|O′ )P (O′ ). (2) O′ Equation 2 captures the basic intuition that the distribution of features for o130 is expected to be consistent with the distribution observed for previous animals. For example, if o130 is known to ﬂy then the trees with high posterior probability P (O′ |D) will be those where o130 is near other ﬂying creatures (Figure 1a), and since these creatures have wings Equation 2 predicts that o130 probably also has wings. As this example suggests, model O captures dependency relationships between features implicitly, and therefore stands in contrast to models like F that rely on explicit representations of relationships between features. A graphical model over features Model F is an undirected graphical model deﬁned over features. The graph shown in Figure 2b was created by identifying pairs where one feature depends directly on another. The author and a research assistant both independently identiﬁed candidate sets of pairwise dependencies, and Figure 2b was created by merging these sets and reaching agreement about how to handle any discrepancies. As previous researchers have suggested [13, 15], feature dependencies can capture several kinds of relationships. For example, wings enable ﬂying, living in the sea leads to eating ﬁsh, and having no legs rules out jumping far. We work with an undirected graph because some pairs of features depend on each other but there is no clear direction of causal inﬂuence. For example, there is clearly a dependency relationship between being nocturnal and seeing in the dark, but no obvious sense in which one of these features causes the other. We assume that the rows of the object-feature matrix D are generated independently from an undirected graphical model F deﬁned over the feature structure in Figure 2b: P (oi |F). P (D|F) = i Model F includes potential functions for each node and for each edge in the graph. These potentials were learned from matrix D using the UGM toolbox for undirected graphical models [19]. The learned potentials capture both positive and negative relationships: for example, animals that live in the sea tend to eat ﬁsh, and tend not to eat berries. Some pairs of feature values never occur together in matrix D (there are no creatures that ﬂy but do not have wings). We therefore chose to compute maximum a posteriori values of the potential functions rather than maximum likelihood values, and used a diffuse Gaussian prior with a variance of 100 on the entries in each potential. After learning the potentials for model F, we can make predictions about a new object o130 using the distribution P (o130 |F). For example, if o130 is known to ﬂy (Figure 1a), model F predicts that o130 probably has wings because the learned potentials capture a positive dependency between ﬂying and having wings. Combining object and feature relationships There are two simple ways to combine models O and F in order to develop an approach that incorporates both relationships between features and relationships between objects. The output combination model computes the predictions of both models in isolation, then combines these predictions using a weighted sum. The resulting model is similar to a mixture-of-experts model, and to avoid free parameters we use a mixing weight of 0.5. The structure combination model combines the graph structures used by the two models and relies on a set of potentials deﬁned over the resulting graph product. An example of a graph product is shown in Figure 1b, and the potential functions for this graph are inherited from the component models in the natural way. Kemp et al. [11] use a similar approach to combine a functional causal model with an object model O, but note that our structure combination model uses an undirected model F rather than a functional causal model over features. Both combination models capture the intuition that inductive inferences rely on relationships between features and relationships between objects. The output combination model has the virtue of 4 simplicity, and the structure combination model is appealing because it relies on a single integrated representation that captures both relationships between features and relationships between objects. To preview our results, our data suggest that the combination models perform better overall than either O or F in isolation, and that both combination models perform about equally well. Exemplar models We will compare the family of graphical models already described with a family of exemplar models. The key difference between these model families is that the exemplar models do not rely on explicit representations of relationships between objects and relationships between features. Comparing the model families can therefore help to establish whether human inferences rely on representations of this sort. Consider ﬁrst a problem where a reasoner must predict whether object o130 has feature k after observing that it has feature i. An exemplar model addresses the problem by retrieving all previouslyobserved objects with feature i and computing the proportion that have feature k: P (ok = 1|oi = 1) = |f k & f i | |f i | (3) where |f k | is the number of objects in matrix D that have feature k, and |f k & f i | is the number that have both feature k and feature i. Note that we have streamlined our notation by using ok instead of o130 to refer to the kth feature value for object o130 . k Suppose now that the reasoner observes that object o130 has features i and j. The natural generalization of Equation 3 is: P (ok = 1|oi = 1, oj = 1) = |f k & f i & f j | |f i & f j | (4) Because we focus on chimeras, |f i & f j | = 0 and Equation 4 is not well deﬁned. We therefore evaluate an exemplar model that computes predictions for the two observed features separately then computes the weighted sum of these predictions: P (ok = 1|oi = 1, oj = 1) = wi |f k & f i | |f k & f j | + wj . i| |f |f j | (5) where the weights wi and wj must sum to one. We consider four ways in which the weights could be set. The ﬁrst strategy sets wi = wj = 0.5. The second strategy sets wi ∝ |f i |, and is consistent with an approach where the reasoner retrieves all exemplars in D that are most similar to the novel animal and reports the proportion of these exemplars that have feature k. The third strategy sets wi ∝ |f1i | , and captures the idea that features should be weighted by their distinctiveness [20]. The ﬁnal strategy sets weights according to the coherence of each feature [21]. A feature is coherent if objects with that feature tend to resemble each other overall, and we deﬁne the coherence of feature i as the expected Jaccard similarity between two randomly chosen objects from matrix D that both have feature i. Note that the ﬁnal three strategies are all consistent with previous proposals from the psychological literature, and each one might be expected to perform well. Because exemplar models and prototype models are often compared, it is natural to consider a prototype model [22] as an additional baseline. A standard prototype model would partition the 129 animals into categories and would use summary statistics for these categories to make predictions about the novel animal o130 . We will not evaluate this model because it corresponds to a coarser version of model O, which organizes the animals into a hierarchy of categories. The key characteristic shared by both models is that they explicitly capture relationships between objects but not features. 2 Experiment 1: Chimeras Our ﬁrst experiment explores how people make inferences about chimeras, or novel animals with features that have not previously been observed to co-occur. Inferences about chimeras raise challenges for exemplar models, and therefore help to establish whether humans rely on explicit representations of relationships between features. Each argument can be represented as f i , f j → f k 5 exemplar r = 0.42 7 feature F exemplar (wi = |f i |) (wi = 0.5) r = 0.44 7 object O r = 0.69 7 output combination r = 0.31 7 structure combination r = 0.59 7 r = 0.60 7 5 5 5 5 5 3 3 3 3 3 3 all 5 1 1 0 1 r = 0.06 7 conflict 0.5 1 1 0 0.5 1 r = 0.71 7 1 0 0.5 1 r = −0.02 7 1 0 0.5 1 r = 0.49 7 0 5 5 5 5 3 3 3 3 1 5 3 0.5 r = 0.57 7 5 3 1 0 0.5 1 r = 0.51 7 edge 0.5 r = 0.17 7 1 1 0 0.5 1 r = 0.64 7 1 0 0.5 1 r = 0.83 7 1 0 0.5 1 r = 0.45 7 1 0 0.5 1 r = 0.76 7 0 5 5 5 5 3 3 3 3 1 5 3 0.5 r = 0.79 7 5 3 1 1 0 0.5 1 r = 0.26 7 other 1 0 1 0 0.5 1 r = 0.25 7 1 0 0.5 1 r = 0.19 7 1 0 0.5 1 r = 0.25 7 1 0 0.5 1 r = 0.24 7 0 7 5 5 5 5 5 3 3 3 3 1 5 3 0.5 r = 0.33 3 1 1 0 0.5 1 1 0 0.5 1 1 0 0.5 1 1 0 0.5 1 1 0 0.5 1 0 0.5 1 Figure 3: Argument ratings for Experiment 1 plotted against the predictions of six models. The y-axis in each panel shows human ratings on a seven point scale, and the x-axis shows probabilities according to one of the models. Correlation coefﬁcients are shown for each plot. where f i and f k are the premises (e.g. “has no legs” and “can ﬂy”) and f k is the conclusion (e.g. “has wings”). We are especially interested in conﬂict cases where the premises f i and f j lead to opposite conclusions when taken individually: for example, most animals with no legs do not have wings, but most animals that ﬂy do have wings. Our models that incorporate feature structure F can resolve this conﬂict since F includes a dependency between “wings” and “can ﬂy” but not between “wings” and “has no legs.” Our models that do not include F cannot resolve the conﬂict and predict that humans will be uncertain about whether the novel animal has wings. Materials. The object-feature matrix D includes 447 feature pairs {f i , f j } such that none of the 129 animals has both f i and f j . We selected 40 pairs (see the supporting material) and created 400 arguments in total by choosing 10 conclusion features for each pair. The arguments can be assigned to three categories. Conﬂict cases are arguments f i , f j → f k such that the single-premise arguments f i → f k and f j → f k lead to incompatible predictions. For our purposes, two singlepremise arguments with the same conclusion are deemed incompatible if one leads to a probability greater than 0.9 according to Equation 3, and the other leads to a probability less than 0.1. Edge cases are arguments f i , f j → f k such that the feature network in Figure 2b includes an edge between f k and either f i or f j . Note that some arguments are both conﬂict cases and edge cases. All arguments that do not fall into either one of these categories will be referred to as other cases. The 400 arguments for the experiment include 154 conﬂict cases, 153 edge cases, and 120 other cases. 34 arguments are both conﬂict cases and edge cases. We chose these arguments based on three criteria. First, we avoided premise pairs that did not co-occur in matrix D but that co-occur in familiar animals that do not belong to D. For example, “is pink” and “has wings” do not co-occur in D but “ﬂamingo” is a familiar animal that has both features. Second, we avoided premise pairs that speciﬁed two different numbers of legs—for example, {“has four legs,” “has six legs”}. Finally, we aimed to include roughly equal numbers of conﬂict cases, edge cases, and other cases. Method. 16 undergraduates participated for course credit. The experiment was carried out using a custom-built computer interface, and one argument was presented on screen at a time. Participants 6 rated the probability of the conclusion on seven point scale where the endpoints were labeled “very unlikely” and “very likely.” The ten arguments for each pair of premises were presented in a block, but the order of these blocks and the order of the arguments within these blocks were randomized across participants. Results. Figure 3 shows average human judgments plotted against the predictions of six models. The plots in the ﬁrst row include all 400 arguments in the experiment, and the remaining rows show results for conﬂict cases, edge cases, and other cases. The previous section described four exemplar models, and the two shown in Figure 3 are the best performers overall. Even though the graphical models include more numerical parameters than the exemplar models, recall that these parameters are learned from matrix D rather than ﬁt to the experimental data. Matrix D also serves as the basis for the exemplar models, which means that all of the models can be compared on equal terms. The ﬁrst row of Figure 3 suggests that the three models which include feature structure F perform better than the alternatives. The output combination model is the worst of the three models that incorporate F, and the correlation achieved by this model is signiﬁcantly greater than the correlation achieved by the best exemplar model (p < 0.001, using the Fisher transformation to convert correlation coefﬁcients to z scores). Our data therefore suggest that explicit representations of relationships between features are needed to account for inductive inferences about chimeras. The model that includes the feature structure F alone performs better than the two models that combine F with the object structure O, which may not be surprising since Experiment 1 focuses speciﬁcally on novel animals that do not slot naturally into structure O. Rows two through four suggest that the conﬂict arguments in particular raise challenges for the models which do not include feature structure F. Since these conﬂict cases are arguments f i , f j → f k where f i → f k has strength greater than 0.9 and f j → f k has strength less than 0.1, the ﬁrst exemplar model averages these strengths and assigns an overall strength of around 0.5 to each argument. The second exemplar model is better able to differentiate between the conﬂict arguments, but still performs substantially worse than the three models that include structure F. The exemplar models perform better on the edge arguments, but are outperformed by the models that include F. Finally, all models achieve roughly the same level of performance on the other arguments. Although the feature model F performs best overall, the predictions of this model still leave room for improvement. The two most obvious outliers in the third plot in the top row represent the arguments {is blue, lives in desert → lives in woods} and {is pink, lives in desert → lives in woods}. Our participants sensibly infer that any animal which lives in the desert cannot simultaneously live in the woods. In contrast, the Leuven database indicates that eight of the twelve animals that live in the desert also live in the woods, and the edge in Figure 2b between “lives in the desert” and “lives in the woods” therefore represents a positive dependency relationship according to model F. This discrepancy between model and participants reﬂects the fact that participants made inferences about individual animals but the Leuven database is based on features of animal categories. Note, for example, that any individual animal is unlikely to live in the desert and the woods, but that some animal categories (including snakes, salamanders, and lizards) are found in both environments. 3 Experiment 2: Single-premise arguments Our results so far suggest that inferences about chimeras rely on explicit representations of relationships between features but provide no evidence that relationships between objects are important. It would be a mistake, however, to conclude that relationships between objects play no role in inductive reasoning. Previous studies have used object structures like the example in Figure 2a to account for inferences about novel features [11]—for example, given that alligators have enzyme Y132 in their blood, it seems likely that crocodiles also have this enzyme. Inferences about novel objects can also draw on relationships between objects rather than relationships between features. For example, given that a novel animal has a beak you will probably predict that it has feathers, not because there is any direct dependency between these two features, but because the beaked animals that you know tend to have feathers. Our second experiment explores inferences of this kind. Materials and Method. 32 undergraduates participated for course credit. The task was identical to Experiment 1 with the following exceptions. Each two-premise argument f i , f j → f k from Experiment 1 was converted into two one-premise arguments f i → f k and f j → f k , and these 7 feature F exemplar r = 0.78 7 object O r = 0.54 7 output combination r = 0.75 7 structure combination r = 0.75 7 all 5 5 5 5 5 3 3 3 3 3 1 1 0 edge 0.5 1 r = 0.87 7 1 0 0.5 1 r = 0.87 7 1 0 0.5 1 r = 0.84 7 1 0 0.5 1 r = 0.86 7 0 5 5 5 3 3 3 1 5 3 0.5 r = 0.85 7 5 3 1 1 0 0.5 1 r = 0.79 7 other r = 0.77 7 1 0 0.5 1 r = 0.21 7 1 0 0.5 1 r = 0.74 7 1 0 0.5 1 r = 0.66 7 0 5 5 5 5 3 3 3 3 1 r = 0.73 7 5 0.5 3 1 1 0 0.5 1 1 0 0.5 1 1 0 0.5 1 1 0 0.5 1 0 0.5 1 Figure 4: Argument ratings and model predictions for Experiment 2. one-premise arguments were randomly assigned to two sets. 16 participants rated the 400 arguments in the ﬁrst set, and the other 16 rated the 400 arguments in the second set. Results. Figure 4 shows average human ratings for the 800 arguments plotted against the predictions of ﬁve models. Unlike Figure 3, Figure 4 includes a single exemplar model since there is no need to consider different feature weightings in this case. Unlike Experiment 1, the feature model F performs worse than the other alternatives (p < 0.001 in all cases). Not surprisingly, this model performs relatively well for edge cases f j → f k where f j and f k are linked in Figure 2b, but the ﬁnal row shows that the model performs poorly across the remaining set of arguments. Taken together, Experiments 1 and 2 suggest that relationships between objects and relationships between features are both needed to account for human inferences. Experiment 1 rules out an exemplar approach but models that combine graph structures over objects and features perform relatively well in both experiments. We considered two methods for combining these structures and both performed equally well. Combining the knowledge captured by these structures appears to be important, and future studies can explore in detail how humans achieve this combination. 4 Conclusion This paper proposed that graphical models are useful for capturing knowledge about animals and their features and showed that a graphical model over features can account for human inferences about chimeras. A family of exemplar models and a graphical model deﬁned over objects were unable to account for our data, which suggests that humans rely on mental representations that explicitly capture dependency relationships between features. Psychologists have previously used graphical models to capture relationships between features, but our work is the ﬁrst to focus on chimeras and to explore models deﬁned over a large set of familiar features. Although a simple undirected model accounted relatively well for our data, this model is only a starting point. The model incorporates dependency relationships between features, but people know about many speciﬁc kinds of dependencies, including cases where one feature causes, enables, prevents, or is inconsistent with another. An undirected graph with only one class of edges cannot capture this knowledge in full, and richer representations will ultimately be needed in order to provide a more complete account of human reasoning. Acknowledgments I thank Madeleine Clute for assisting with this research. This work was supported in part by the Pittsburgh Life Sciences Greenhouse Opportunity Fund and by NSF grant CDI-0835797. 8 References [1] R. N. Shepard. Towards a universal law of generalization for psychological science. Science, 237:1317– 1323, 1987. [2] J. R. Anderson. The adaptive nature of human categorization. Psychological Review, 98(3):409–429, 1991. [3] E. Heit. A Bayesian analysis of some forms of inductive reasoning. In M. Oaksford and N. Chater, editors, Rational models of cognition, pages 248–274. Oxford University Press, Oxford, 1998. [4] J. B. Tenenbaum and T. L. Grifﬁths. Generalization, similarity, and Bayesian inference. Behavioral and Brain Sciences, 24:629–641, 2001. [5] C. Kemp and J. B. Tenenbaum. Structured statistical models of inductive reasoning. Psychological Review, 116(1):20–58, 2009. [6] D. N. Osherson, E. E. Smith, O. Wilkie, A. Lopez, and E. Shaﬁr. Category-based induction. Psychological Review, 97(2):185–200, 1990. [7] D. J. Navarro. Learning the context of a category. In J. Lafferty, C. K. I. Williams, J. Shawe-Taylor, R.S. Zemel, and A. Culotta, editors, Advances in Neural Information Processing Systems 23, pages 1795–1803. 2010. [8] C. Kemp, T. L. Grifﬁths, S. Stromsten, and J. B. Tenenbaum. Semi-supervised learning with trees. In Advances in Neural Information Processing Systems 16, pages 257–264. MIT Press, Cambridge, MA, 2004. [9] B. Rehder. A causal-model theory of conceptual representation and categorization. Journal of Experimental Psychology: Learning, Memory, and Cognition, 29:1141–1159, 2003. [10] B. Rehder and R. Burnett. Feature inference and the causal structure of categories. Cognitive Psychology, 50:264–314, 2005. [11] C. Kemp, P. Shafto, and J. B. Tenenbaum. An integrated account of generalization across objects and features. Cognitive Psychology, in press. [12] S. E. Barrett, H. Abdi, G. L. Murphy, and J. McCarthy Gallagher. Theory-based correlations and their role in children’s concepts. Child Development, 64:1595–1616, 1993. [13] S. A. Sloman, B. C. Love, and W. Ahn. Feature centrality and conceptual coherence. Cognitive Science, 22(2):189–228, 1998. [14] D. Yarlett and M. Ramscar. A quantitative model of counterfactual reasoning. In T. G. Dietterich, S. Becker, and Z. Ghahramani, editors, Advances in Neural Information Processing Systems 14, pages 123–130. MIT Press, Cambridge, MA, 2002. [15] W. Ahn, J. K. Marsh, C. C. Luhmann, and K. Lee. Effect of theory-based feature correlations on typicality judgments. Memory and Cognition, 30(1):107–118, 2002. [16] D. C. Meehan C. McNorgan, R. A. Kotack and K. McRae. Feature-feature causal relations and statistical co-occurrences in object concepts. Memory and Cognition, 35(3):418–431, 2007. [17] S. De Deyne, S. Verheyen, E. Ameel, W. Vanpaemel, M. J. Dry, W. Voorspoels, and G. Storms. Exemplar by feature applicability matrices and other Dutch normative data for semantic concepts. Behavior Research Methods, 40(4):1030–1048, 2008. [18] J. P. Huelsenbeck and F. Ronquist. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics, 17(8):754–755, 2001. [19] M. Schmidt. UGM: A Matlab toolbox for probabilistic undirected graphical models. 2007. Available at http://people.cs.ubc.ca/∼schmidtm/Software/UGM.html. [20] L. J. Nelson and D. T. Miller. The distinctiveness effect in social categorization: you are what makes you unusual. Psychological Science, 6:246–249, 1995. [21] A. L. Patalano, S. Chin-Parker, and B. H. Ross. The importance of being coherent: category coherence, cross-classiﬁcation and reasoning. Journal of memory and language, 54:407–424, 2006. [22] S. K. Reed. Pattern recognition and categorization. Cognitive Psychology, 3:393–407, 1972. 9

same-paper 2 0.89858389 193 nips-2011-Object Detection with Grammar Models

Author: Ross B. Girshick, Pedro F. Felzenszwalb, David A. McAllester

3 0.88299805 39 nips-2011-Approximating Semidefinite Programs in Sublinear Time

Author: Dan Garber, Elad Hazan

Abstract: In recent years semideﬁnite optimization has become a tool of major importance in various optimization and machine learning problems. In many of these problems the amount of data in practice is so large that there is a constant need for faster algorithms. In this work we present the ﬁrst sublinear time approximation algorithm for semideﬁnite programs which we believe may be useful for such problems in which the size of data may cause even linear time algorithms to have prohibitive running times in practice. We present the algorithm and its analysis alongside with some theoretical lower bounds and an improved algorithm for the special problem of supervised learning of a distance metric. 1

4 0.85632724 213 nips-2011-Phase transition in the family of p-resistances

Author: Morteza Alamgir, Ulrike V. Luxburg

Abstract: We study the family of p-resistances on graphs for p 1. This family generalizes the standard resistance distance. We prove that for any ﬁxed graph, for p = 1 the p-resistance coincides with the shortest path distance, for p = 2 it coincides with the standard resistance distance, and for p ! 1 it converges to the inverse of the minimal s-t-cut in the graph. Secondly, we consider the special case of random geometric graphs (such as k-nearest neighbor graphs) when the number n of vertices in the graph tends to inﬁnity. We prove that an interesting phase transition takes place. There exist two critical thresholds p⇤ and p⇤⇤ such that if p < p⇤ , then the p-resistance depends on meaningful global properties of the graph, whereas if p > p⇤⇤ , it only depends on trivial local quantities and does not convey any useful information. We can explicitly compute the critical values: p⇤ = 1 + 1/(d 1) and p⇤⇤ = 1 + 1/(d 2) where d is the dimension of the underlying space (we believe that the fact that there is a small gap between p⇤ and p⇤⇤ is an artifact of our proofs). We also relate our ﬁndings to Laplacian regularization and suggest to use q-Laplacians as regularizers, where q satisﬁes 1/p⇤ + 1/q = 1. 1

5 0.79766339 279 nips-2011-Target Neighbor Consistent Feature Weighting for Nearest Neighbor Classification

Author: Ichiro Takeuchi, Masashi Sugiyama

Abstract: We consider feature selection and weighting for nearest neighbor classiﬁers. A technical challenge in this scenario is how to cope with discrete update of nearest neighbors when the feature space metric is changed during the learning process. This issue, called the target neighbor change, was not properly addressed in the existing feature weighting and metric learning literature. In this paper, we propose a novel feature weighting algorithm that can exactly and efﬁciently keep track of the correct target neighbors via sequential quadratic programming. To the best of our knowledge, this is the ﬁrst algorithm that guarantees the consistency between target neighbors and the feature space metric. We further show that the proposed algorithm can be naturally combined with regularization path tracking, allowing computationally efﬁcient selection of the regularization parameter. We demonstrate the effectiveness of the proposed algorithm through experiments. 1

6 0.77474922 139 nips-2011-Kernel Bayes' Rule

7 0.68114769 127 nips-2011-Image Parsing with Stochastic Scene Grammar

8 0.58227819 64 nips-2011-Convergent Bounds on the Euclidean Distance

9 0.55789894 242 nips-2011-See the Tree Through the Lines: The Shazoo Algorithm

10 0.54991591 154 nips-2011-Learning person-object interactions for action recognition in still images

11 0.5380373 166 nips-2011-Maximal Cliques that Satisfy Hard Constraints with Application to Deformable Object Model Learning

12 0.52458549 75 nips-2011-Dynamical segmentation of single trials from population neural data

13 0.5183391 168 nips-2011-Maximum Margin Multi-Instance Learning

14 0.51115167 233 nips-2011-Rapid Deformable Object Detection using Dual-Tree Branch-and-Bound

15 0.51096416 204 nips-2011-Online Learning: Stochastic, Constrained, and Smoothed Adversaries

16 0.51056093 231 nips-2011-Randomized Algorithms for Comparison-based Search

17 0.49276665 150 nips-2011-Learning a Distance Metric from a Network

18 0.48914003 303 nips-2011-Video Annotation and Tracking with Active Learning

19 0.47962755 103 nips-2011-Generalization Bounds and Consistency for Latent Structural Probit and Ramp Loss

20 0.46786195 27 nips-2011-Advice Refinement in Knowledge-Based SVMs