Paying Attention to Descriptions Generated by Image Captioning Models

To bridge the gap between humans and machines in image understanding and describing, we need further insight into how people describe a perceived scene. In this paper, we study the agreement between bottom-up saliency-based visual attention and object referrals in scene description constructs. We investigate the properties of human-written descriptions and machine-generated ones. We then propose a saliency-boosted image captioning model in order to investigate benefits from low-level cues in language models. We learn that (1) humans mention more salient objects earlier than less salient ones in their descriptions, (2) the better a captioning model performs, the better attention agreement it has with human descriptions, (3) the proposed saliencyboosted model, compared to its baseline form, does not improve significantly on the MS COCO database, indicating explicit bottom-up boosting does not help when the task is well learnt and tuned on a data, (4) a better generalization is, however, observed for the saliency-boosted model on unseen data.

[1]  Ali Borji,et al.  Reconciling Saliency and Object Center-Bias Hypotheses in Explaining Free-Viewing Fixations , 2015, IEEE Transactions on Neural Networks and Learning Systems.

[2]  Ali Borji,et al.  Analysis of Scores, Datasets, and Models in Visual Saliency Prediction , 2013, 2013 IEEE International Conference on Computer Vision.

[3]  W. Einhäuser,et al.  Fixations on objects in natural scenes: dissociating importance from salience , 2013, Front. Psychol..

[4]  Yifan Peng,et al.  Exploring the role of gaze behavior and object detection in scene understanding , 2013, Front. Psychol..

[5]  F. Pulvermüller,et al.  Walking or Talking?: Behavioral and Neurophysiological Correlates of Action Verb Processing , 2001, Brain and Language.

[6]  Geoffrey Zweig,et al.  From captions to visual concepts and back , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[7]  Jana Holsanova.  How we focus attention in picture viewing, picture description, and during mental imagery , 2009 .

[8]  Michael A. Arbib,et al.  Action to Language via the Mirror Neuron System: Attention and the minimal subscene , 2006 .

[9]  C. Lawrence Zitnick,et al.  CIDEr: Consensus-based image description evaluation , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[10]  Alon Lavie,et al.  Meteor Universal: Language Specific Translation Evaluation for Any Target Language , 2014, WMT@ACL.

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

[12]  S. N. Sridhar,et al.  Models of Sentence Production , 1988 .

[13]  Mihai Surdeanu,et al.  The Stanford CoreNLP Natural Language Processing Toolkit , 2014, ACL.

[14]  Jeffrey Dean,et al.  Efficient Estimation of Word Representations in Vector Space , 2013, ICLR.

[15]  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.

[16]  Qi Zhao,et al.  Learning saliency-based visual attention: A review , 2013, Signal Process..

[17]  Jorma Laaksonen,et al.  Video captioning with recurrent networks based on frame- and video-level features and visual content classification , 2015, ArXiv.

[18]  Desmond Elliott,et al.  Describing Images using Inferred Visual Dependency Representations , 2015, ACL.

[19]  Frank Keller,et al.  Image Description using Visual Dependency Representations , 2013, EMNLP.

[20]  A. Meyer.  The use of eye tracking in studies of sentence generation , 2004 .

[21]  Jorma Laaksonen,et al.  Saliency Revisited: Analysis of Mouse Movements Versus Fixations , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[22]  Peter Young,et al.  From image descriptions to visual denotations: New similarity metrics for semantic inference over event descriptions , 2014, TACL.

[23]  Bernt Schiele,et al.  The Long-Short Story of Movie Description , 2015, GCPR.

[24]  Yoshua Bengio,et al.  Show, Attend and Tell: Neural Image Caption Generation with Visual Attention , 2015, ICML.

[25]  Fei-Fei Li,et al.  Deep visual-semantic alignments for generating image descriptions , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[26]  Daniel Dominic Sleator,et al.  Parsing English with a Link Grammar , 1995, IWPT.

[27]  Karl Stratos,et al.  Understanding and predicting importance in images , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[28]  Chin-Yew Lin,et al.  ROUGE: A Package for Automatic Evaluation of Summaries , 2004, ACL 2004.

[29]  Luc Van Gool,et al.  The Pascal Visual Object Classes Challenge: A Retrospective , 2014, International Journal of Computer Vision.

[30]  Xinlei Chen,et al.  Microsoft COCO Captions: Data Collection and Evaluation Server , 2015, ArXiv.

[31]  M. Elsner,et al.  Giving Good Directions: Order of Mention Reflects Visual Salience , 2015, Front. Psychol..

[32]  Cordelia Schmid,et al.  Areas of Attention for Image Captioning , 2016, 2017 IEEE International Conference on Computer Vision (ICCV).

[33]  Nazli Ikizler-Cinbis,et al.  Automatic Description Generation from Images: A Survey of Models, Datasets, and Evaluation Measures (Extended Abstract) , 2017, IJCAI.

[34]  Pietro Perona,et al.  Microsoft COCO: Common Objects in Context , 2014, ECCV.

[35]  Kate Saenko,et al.  Top-Down Visual Saliency Guided by Captions , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[36]  Armand Joulin,et al.  Deep Fragment Embeddings for Bidirectional Image Sentence Mapping , 2014, NIPS.

[37]  Zenzi M. Griffin,et al.  Observing the what and when of language production for different age groups by monitoring speakers’ eye movements , 2006, Brain and Language.

[38]  Vicente Ordonez,et al.  Im2Text: Describing Images Using 1 Million Captioned Photographs , 2011, NIPS.

[39]  D. E. Irwin,et al.  Minding the clock , 2003 .

[40]  Margaret Mitchell,et al.  VQA: Visual Question Answering , 2015, International Journal of Computer Vision.

[41]  Peter Young,et al.  Framing Image Description as a Ranking Task: Data, Models and Evaluation Metrics , 2013, J. Artif. Intell. Res..

[42]  Jorma Laaksonen,et al.  Exploiting inter-image similarity and ensemble of extreme learners for fixation prediction using deep features , 2016, Neurocomputing.

[43]  Asha Iyer,et al.  Components of bottom-up gaze allocation in natural images , 2005, Vision Research.

[44]  Yifan Peng,et al.  Studying Relationships between Human Gaze, Description, and Computer Vision , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[45]  Zenzi M. Griffin,et al.  What the Eyes Say About Speaking , 2000, Psychological science.

[46]  Subhransu Maji,et al.  Deep filter banks for texture recognition and segmentation , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[47]  Zenzi M. Griffin,et al.  Why Look? Reasons for Eye Movements Related to Language Production. , 2004 .

[48]  M. Arbib.  Action to language via the mirror neuron system , 2006 .

[49]  J. Charles.  Cognition and Sentence Production: A Cross-Linguistic Study. , 1989 .

[50]  Mario Fritz,et al.  Ask Your Neurons: A Neural-Based Approach to Answering Questions about Images , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[51]  Cyrus Rashtchian,et al.  Collecting Image Annotations Using Amazon’s Mechanical Turk , 2010, Mturk@HLT-NAACL.

[52]  Michelle R. Greene.  Statistics of high-level scene context , 2013, Front. Psychol..

[53]  Yusuke Sugano,et al.  Seeing with Humans: Gaze-Assisted Neural Image Captioning , 2016, ArXiv.

[54]  Qi Zhao,et al.  SALICON: Saliency in Context , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[55]  Frédo Durand,et al.  Learning to predict where humans look , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[56]  Samy Bengio,et al.  Show and tell: A neural image caption generator , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[57]  Jorma Laaksonen,et al.  Exploiting Scene Context for Image Captioning , 2016, iV&L-MM@MM.

[58]  Salim Roukos,et al.  Bleu: a Method for Automatic Evaluation of Machine Translation , 2002, ACL.

[59]  Atsuto Maki,et al.  From generic to specific deep representations for visual recognition , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).

[60]  Cristian Sminchisescu,et al.  Actions in the Eye: Dynamic Gaze Datasets and Learnt Saliency Models for Visual Recognition , 2013, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[61]  Cyrus Rashtchian,et al.  Every Picture Tells a Story: Generating Sentences from Images , 2010, ECCV.

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

[63]  Trevor Darrell,et al.  Long-term recurrent convolutional networks for visual recognition and description , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[64]  Geoffrey Zweig,et al.  Language Models for Image Captioning: The Quirks and What Works , 2015, ACL.

[65]  Chee Kheong Siew,et al.  Extreme learning machine: Theory and applications , 2006, Neurocomputing.

[66]  Pietro Perona,et al.  Graph-Based Visual Saliency , 2006, NIPS.

[67]  W. Levelt,et al.  Viewing and naming objects: eye movements during noun phrase production , 1998, Cognition.

[68]  C. Lawrence Zitnick,et al.  Adopting Abstract Images for Semantic Scene Understanding , 2016, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[69]  Yejin Choi,et al.  Composing Simple Image Descriptions using Web-scale N-grams , 2011, CoNLL.