Dataset bias: A case study for visual question answering

We examine the issue of bias in datasets designed to train visual question answering (VQA) algorithms. These datasets include a collection of natural language questions about images (aka ‐ visual questions). We consider three popular datasets that are captured by people with sight, people who are blind, and generated by computers. We first demonstrate that machine learning algorithms can be trained to recognize each dataset's bias, and so determine the source of a novel visual question. We then discuss potential risks and benefits of biased VQA datasets and corresponding machine learning algorithms that can identify the source of a visual question; e.g., whether it comes from a person with sight, a person who is blind, or bot (aka ‐ computer). Our ultimate aim is to inspire the development of more inclusive VQA systems.

[1]  Li Fei-Fei,et al.  ImageNet: A large-scale hierarchical image database , 2009, CVPR.

[2]  Alex Pentland,et al.  Fair, Transparent, and Accountable Algorithmic Decision-making Processes , 2017, Philosophy & Technology.

[3]  Leif Azzopardi,et al.  The impact of fielding on retrieval performance and bias , 2018 .

[4]  Christopher Kanan,et al.  Visual question answering: Datasets, algorithms, and future challenges , 2016, Comput. Vis. Image Underst..

[5]  Toniann Pitassi,et al.  Fairness through awareness , 2011, ITCS '12.

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

[7]  Jiebo Luo,et al.  VizWiz Grand Challenge: Answering Visual Questions from Blind People , 2018, 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition.

[8]  Ricardo Baeza-Yates,et al.  FA*IR: A Fair Top-k Ranking Algorithm , 2017, CIKM.

[9]  Timnit Gebru,et al.  Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification , 2018, FAT.

[10]  Jieyu Zhao,et al.  Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraints , 2017, EMNLP.

[11]  Michael McGill,et al.  Introduction to Modern Information Retrieval , 1983 .

[12]  Li Fei-Fei,et al.  CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[13]  Dawn O. Braithwaite,et al.  “Just How Much Did That Wheelchair Cost?”: Management of Privacy Boundaries by Persons with Disabilities , 1991 .

[14]  Allan Jabri,et al.  Revisiting Visual Question Answering Baselines , 2016, ECCV.

[15]  Boqing Gong,et al.  Overcoming Dataset Bias : An Unsupervised Domain Adaptation Approach , 2012 .

[16]  Adriana Kovashka,et al.  Seeing Behind the Camera: Identifying the Authorship of a Photograph , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[17]  Michael A. Katell,et al.  Algorithmic reputation and information equity , 2018, ASIST.

[18]  Antonio Torralba,et al.  LabelMe: A Database and Web-Based Tool for Image Annotation , 2008, International Journal of Computer Vision.

[19]  Foad Hamidi,et al.  Who Should Have Access to my Pointing Data?: Privacy Tradeoffs of Adaptive Assistive Technologies , 2018, ASSETS.

[20]  Qi Wu,et al.  Visual question answering: A survey of methods and datasets , 2016, Comput. Vis. Image Underst..

[21]  Martin Dillon,et al.  Introduction to modern information retrieval: G. Salton and M. McGill. McGraw-Hill, New York (1983). xv + 448 pp., $32.95 ISBN 0-07-054484-0 , 1983 .

[22]  Xinlei Chen,et al.  Pythia v0.1: the Winning Entry to the VQA Challenge 2018 , 2018, ArXiv.

[23]  Dumitru Erhan,et al.  Show and Tell: Lessons Learned from the 2015 MSCOCO Image Captioning Challenge , 2016, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[24]  G. Griffin,et al.  Caltech-256 Object Category Dataset , 2007 .

[25]  Alexei A. Efros,et al.  Unbiased look at dataset bias , 2011, CVPR 2011.

[26]  Luc Van Gool,et al.  The Pascal Visual Object Classes (VOC) Challenge , 2010, International Journal of Computer Vision.

[27]  Rachel K. E. Bellamy,et al.  AI Fairness 360: An Extensible Toolkit for Detecting, Understanding, and Mitigating Unwanted Algorithmic Bias , 2018, ArXiv.

[28]  Jakub Hajič,et al.  Visual Question Answering , 2022, International Journal of Advanced Research in Science, Communication and Technology.

[29]  Alexandra Chouldechova,et al.  Fair prediction with disparate impact: A study of bias in recidivism prediction instruments , 2016, Big Data.

[30]  Alexei A. Efros,et al.  Undoing the Damage of Dataset Bias , 2012, ECCV.

[31]  Erin Brady,et al.  Visual challenges in the everyday lives of blind people , 2013, CHI.

[32]  Vitaly Shmatikov,et al.  Robust De-anonymization of Large Sparse Datasets , 2008, 2008 IEEE Symposium on Security and Privacy (sp 2008).

[33]  Jiasen Lu,et al.  Hierarchical Question-Image Co-Attention for Visual Question Answering , 2016, NIPS.

[34]  Filippo Menczer,et al.  Quantifying Biases in Online Information Exposure , 2018, J. Assoc. Inf. Sci. Technol..

[35]  Hai Zhuge,et al.  Exploiting heterogeneous scientific literature networks to combat ranking bias: Evidence from the computational linguistics area , 2016, J. Assoc. Inf. Sci. Technol..

[36]  Reuben Binns,et al.  Fairness in Machine Learning: Lessons from Political Philosophy , 2017, FAT.

[37]  Christine T. Wolf,et al.  Beyond fairness, accountability, and transparency in the ethics of algorithms: Contributions and perspectives from LIS , 2018, ASIST.

[38]  Rob Miller,et al.  VizWiz: nearly real-time answers to visual questions , 2010, UIST.

[39]  Barbara Caputo,et al.  A Deeper Look at Dataset Bias , 2015, Domain Adaptation in Computer Vision Applications.

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

[41]  Niall McLaughlin,et al.  Data-augmentation for reducing dataset bias in person re-identification , 2015, 2015 12th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS).