Network Dissection: Quantifying Interpretability of Deep Visual Representations

We propose a general framework called Network Dissection for quantifying the interpretability of latent representations of CNNs by evaluating the alignment between individual hidden units and a set of semantic concepts. Given any CNN model, the proposed method draws on a data set of concepts to score the semantics of hidden units at each intermediate convolutional layer. The units with semantics are labeled across a broad range of visual concepts including objects, parts, scenes, textures, materials, and colors. We use the proposed method to test the hypothesis that interpretability is an axis-independent property of the representation space, then we apply the method to compare the latent representations of various networks when trained to solve different classification problems. We further analyze the effect of training iterations, compare networks trained with different initializations, and measure the effect of dropout and batch normalization on the interpretability of deep visual representations. We demonstrate that the proposed method can shed light on characteristics of CNN models and training methods that go beyond measurements of their discriminative power.

[1]  C. Koch,et al.  Invariant visual representation by single neurons in the human brain , 2005, Nature.

[2]  Cordelia Schmid,et al.  Learning Color Names for Real-World Applications , 2009, IEEE Transactions on Image Processing.

[3]  Leonidas J. Guibas,et al.  Human action recognition by learning bases of action attributes and parts , 2011, 2011 International Conference on Computer Vision.

[4]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[5]  Pascal Vincent,et al.  Representation Learning: A Review and New Perspectives , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[6]  Yoshua Bengio,et al.  How transferable are features in deep neural networks? , 2014, NIPS.

[7]  Iasonas Kokkinos,et al.  Describing Textures in the Wild , 2013, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[8]  Rob Fergus,et al.  Visualizing and Understanding Convolutional Networks , 2013, ECCV.

[9]  Sanja Fidler,et al.  The Role of Context for Object Detection and Semantic Segmentation in the Wild , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[10]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[11]  Noah Snavely,et al.  Intrinsic images in the wild , 2014, ACM Trans. Graph..

[12]  Stefan Carlsson,et al.  CNN Features Off-the-Shelf: An Astounding Baseline for Recognition , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops.

[13]  Bolei Zhou,et al.  Learning Deep Features for Scene Recognition using Places Database , 2014, NIPS.

[14]  Sanja Fidler,et al.  Detect What You Can: Detecting and Representing Objects Using Holistic Models and Body Parts , 2014, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[15]  Joan Bruna,et al.  Intriguing properties of neural networks , 2013, ICLR.

[16]  Andrew Zisserman,et al.  Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps , 2013, ICLR.

[17]  Jitendra Malik,et al.  Analyzing the Performance of Multilayer Neural Networks for Object Recognition , 2014, ECCV.

[18]  Sergey Ioffe,et al.  Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift , 2015, ICML.

[19]  Andrea Vedaldi,et al.  Understanding deep image representations by inverting them , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[20]  Nitish Srivastava Unsupervised Learning of Visual Representations using Videos , 2015 .

[21]  Bolei Zhou,et al.  Object Detectors Emerge in Deep Scene CNNs , 2014, ICLR.

[22]  Kristen Grauman,et al.  Learning Image Representations Tied to Ego-Motion , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[23]  Hod Lipson,et al.  Convergent Learning: Do different neural networks learn the same representations? , 2015, FE@NIPS.

[24]  Jitendra Malik,et al.  Learning to See by Moving , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[25]  Dumitru Erhan,et al.  Going deeper with convolutions , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[26]  Michael S. Bernstein,et al.  ImageNet Large Scale Visual Recognition Challenge , 2014, International Journal of Computer Vision.

[27]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[28]  Alexei A. Efros,et al.  Unsupervised Visual Representation Learning by Context Prediction , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[29]  Nikos Komodakis,et al.  Wide Residual Networks , 2016, BMVC.

[30]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[31]  Paolo Favaro,et al.  Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles , 2016, ECCV.

[32]  Martial Hebert,et al.  Shuffle and Learn: Unsupervised Learning Using Temporal Order Verification , 2016, ECCV.

[33]  Alexei A. Efros,et al.  Colorful Image Colorization , 2016, ECCV.

[34]  Andrew Owens,et al.  Ambient Sound Provides Supervision for Visual Learning , 2016, ECCV.

[35]  Bolei Zhou,et al.  Places: An Image Database for Deep Scene Understanding , 2016, ArXiv.

[36]  Kristen Grauman,et al.  Object-Centric Representation Learning from Unlabeled Videos , 2016, ACCV.

[37]  Thomas Brox,et al.  Synthesizing the preferred inputs for neurons in neural networks via deep generator networks , 2016, NIPS.

[38]  Antonio Torralba,et al.  Generating Videos with Scene Dynamics , 2016, NIPS.

[39]  Bolei Zhou,et al.  Scene Parsing through ADE20K Dataset , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[40]  Davide Modolo,et al.  Do Semantic Parts Emerge in Convolutional Neural Networks? , 2016, International Journal of Computer Vision.

[41]  Yoshua Bengio,et al.  Understanding intermediate layers using linear classifier probes , 2016, ICLR.

[42]  Alexei A. Efros,et al.  Split-Brain Autoencoders: Unsupervised Learning by Cross-Channel Prediction , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).