AI

IBM Research AI at CVPR 2019

Share this post:

The annual conference on Computer Vision and Pattern Recognition (CVPR 2019) takes place June 16–20 in Long Beach, CA. IBM Research AI will present technical papers describing our latest results in our quest to give AI systems sight. Our research explores a variety of areas within computer vision and multimedia, with a focus on multimodal perception (jointly modeling vision, sound, and language), video understanding, and “learning more from less” — i.e., learning more accurate models with less data, expertise, and effort.

IBM Research is also proudly sponsoring CVPR 2019 at the Platinum level. And we are particularly excited to once again sponsor the Women in Computer Vision workshop. At our booth #513 in the conference expo, we are featuring interactive demos of our latest computer vision technologies, including a few-shot custom object learning technique deployed in a real-world application for food recognition, a multimodal system for auto-curation of sports highlights (used to produce the official highlights of the Masters 2019 golf tournament, and previous US Open and Wimbledon tennis tournaments), an interactive system for fashion search based on natural language feedback, and much more.

Please stop by the booth and attend our presentations and workshops listed below. See you in Long Beach!

Accepted papers at the main conference

P3SGD: Patient Privacy Preserving SGD for Regularizing Deep CNNs in Pathological Image Classification
Bingzhe Wu, Shiwan Zhao, Guangyu Sun, Xiaolu Zhang, Zhong Su, Caihong Zeng, Zhihong Liu
Tuesday June 18, 10:15–13:00

Geometry-Aware Distillation for Indoor Semantic Segmentation
Jianbo Jiao, Yunchao Wei, Zequn Jie, Honghui Shi, Rynson W.H. Lau, Thomas S. Huang
Tuesday June 18, 15:20–18:00

Unifying Heterogeneous Classifiers with Distillation
Jayakorn Vongkulbhisal, Phongtharin Vinayavekhin, Marco Visentini
Tuesday June 18, 15:20–18:00

RepMet: Representative-Based Metric Learning for Classification and One-shot Object Detection
Leonid Karlinsky, Joseph Shtok, Sivan Harary, Eli Schwartz, Amit Aides, Rogerio Feris, Raja Giryes, Alex M. Bronstein
Wednesday June 19, 10:00–12:45
(Read blog post)

SpotTune: Transfer Learning through Adaptive Fine-Tuning
Yunhui Guo, Honghui Shi, Abhishek Kumar, Kristen Grauman, Tajana Rosing, Rogerio Feris
Wednesday June 19, 10:00–12:45
(Read blog post)

Text2Scene: Generating Compositional Scenes from Textual Descriptions
Fuwen Tan, Song Feng, Vicente Ordonez
Wednesday June 19, 14:47 [Oral]
Wednesday June 19, 15:20–18:00
(Read blog post)

LaSO: Label-Set Operations Networks for Multi-Label Few-Shot Learning
Leonid Karlinsky, Amit Aides, Joseph Shtok, Sivan Harary, Rogerio Feris, Raja Giryes, Alex M. Bronstein
Wednesday June 19, 15:00 [Oral]
Wednesday June 19, 15:20–18:00
(Read blog post)

Bayesian Hierarchical Dynamic Model for Human Action Recognition
Rui Zhao, Wanru Xu, Hui Su, Qiang Ji
Wednesday June 19, 15:20–18:00

Adversarial Semantic Alignment for Improved Image Captions 
Igor Melnyk, Tom Sercu, Pierre L Dognin, Jarret Ross, Youssef Mroueh
Thursday June 20, 10:00–12:45
(Read blog post)

Neuro-Inspired Eye Tracking with Eye Movement Dynamics
Kang Wang, Hui Su, Qiang Ji
Thursday June 20, 10:00–12:45

Transferable AutoML by Model Sharing over Grouped Datasets
Chao Xue, JunChi Yan, RongYan, Stephen Chu, Yonghua Lin and YongGang Hu
Thursday June 20, 10:00–12:45

Additive Adversarial Learning for Unbiased Authentication
Jian Liang, Yuren Cao, Chenbin Zhang, Shiyu Chang (IBM), Kun Bai, Zenglin Xu
Thursday June 20, 15:20–18:00

Generalizing Eye Tracking with Bayesian Adversarial Learning
Kang Wang, Rui Zhao, Hui Su, Qiang Ji
Thursday June 20, 15:20–18:00

Unsupervised Learning of Action Classes with Continuous Temporal Embedding
Anna Kukleva, Hilde Kuehne, Fadime Sener, Jürgen Gall
Thursday June 20, 15:20–18:00

Accepted workshop papers

Baby steps towards few-shot learning with multiple semantics
Eli Schwartz*, Leonid Karlinsky*, Rogerio Feris, Raja Giryes and Alex M. Bronstein
Language and Vision Workshop
Wednesday June 16

The Fashion IQ Dataset: Retrieving Images by Combining Side Information and Relative Natural Language Feedback
Xiaoxiao Guo*, Hui Wu*, Yupeng Gao, Steven Rennie, Rogerio Feris
Visual Question Answering and Dialog Workshop
Monday June 17

Automatic Labeling of Data for Transfer Learning
Parijat Dube, Bishwaranjan Bhattacharjee, Siyu Huo, Patrick Watson, Brian Belgodere, John Kender
Deep Vision Workshop
Sunday June 16

Identifying Interpretable Action Concepts in Deep Networks.
Kandan Ramakrishnan, Mathew Monfort, Barry A McNamara,
Alex Lascelles, Dan Gutfreund, Rogerio Feris, Aude Oliva
Workshop on Explainable AI
Sunday June 16

Grounding Spoken Words in Unlabeled Video
Angie Boggust, Kartik Audhkhasi, Dhiraj Joshi, David Harwath, Samuel Thomas,
Rogerio Feris, Dan Gutfreund, Yang Zhang, Antonio Torralba, Michael Picheny, James Glass.
Sight and Sound Workshop
Monday June 17

Training Deep Neural Networks based on Encrypted Data
Nandakumar, N. Ratha, S. Panakanti, and S. Halevi
Challenges and Opportunity for Privacy and Security (CV-COPS 2019)
Sunday June 16

DeepRing: Protecting Deep Neural Network with Blockchain
A. Goel, A. Agarwal, R. Singh, M. Vatsa, N. Ratha
Blockchain Meets Computer Vision and AI Workshop
Monday June 17

Workshops organized by IBM

Applications and the 1st Learning from Imperfect Data (LID) Challenge
Sunday June 16 @ Hyatt Regency B

Language and Vision
Sunday June 16 @ Seaside 1

Bias Estimation in Face Analytics
Monday June 17 @ Hyatt Seaview B

Blockchain Meets Computer Vision & AI
Monday June 17 @ Hyatt Seaview C

Fourth International Skin Imaging Collaboration (ISIC) Workshop on Skin Image Analysis
Monday June 17 @ Hyatt Seaview B

Multi-Modal Learning from Videos
Monday June 17 @ 201B

Workshops: Invited talks

John R. Smith, “Learning more from Less”
2nd Workshop on Vision with Biased or Scarce Data
Sunday June 16th, 09:50 @ Hyatt Seaview B

John R. Smith, “Multi-modal Perception Understanding at IBM Research”
Workshop on Precognition: Seeing Through the Future
Monday June 17th, 16:00 @ Hyatt Shoreline A

Rogerio Feris, “Is it all Relative? Interactive Fashion Search based on Relative Natural Language Feedback”
Workshop on Understanding Subjective Attributes of Data: Focus on Fashion and Subjective Search
Sunday June 16, 09:20 @ Hyatt Seaview C

Rogerio Feris, “Speeding Up Deep Neural Networks with Adaptive Computation and Efficient Multi-Scale Architectures”
Workshop on EMC^2: Energy Efficient Machine Learning and Cognitive Computing for Embedded Applications
Sunday June 16, 11:40 @ Hyatt Shoreline A

Rogerio Feris, “Learning More from Less: Weak Supervision and Beyond”
Workshop on Weakly Supervised Learning for Real-World Computer Vision Applications and the 1st Learning from Imperfect Data (LID) Challenge
Sunday June 16, 14:00 @ Hyatt Regency B

Accepted Demos

Fashion IQ: Interactive Fashion Retrieval Using Natural Language Feedback
Xiaoxiao Guo, Yupeng Gao, Hui Wu
https://www.spacewu.com/posts/fashion-iq/
Tuesday June 18, 15:20–18:00 Demos (Exhibit Hall)

Accepted Tutorials

Recent Advances in Visual Data Summarization
Rameswar Panda, Ehsan Elhamifar, Amin Karbasi  and Michael Gygli
https://rpand002.github.io/cvpr19_sumt.html
Sunday June 16, 13:30-17:30 @ 203C

Principal RSM and Manager, Computer Vision and Multimedia Department, IBM Research

More AI stories

Pushing the boundaries of convex optimization

Convex optimization problems, which involve the minimization of a convex function over a convex set, can be approximated in theory to any fixed precision in polynomial time. However, practical algorithms are known only for special cases. An important question is whether it is possible to develop algorithms for a broader subset of convex optimization problems that are efficient in both theory and practice.

Continue reading

Making Neural Networks Robust with New Perspectives

IBM researchers have partnered with scientists from MIT, Northeastern University, Boston University and University of Minnesota to publish two papers on novel attacks and defenses for graph neural networks and on a new robust training algorithm called hierarchical random switching at IJCAI 2019.

Continue reading

Improving the Scalability of AI Planning when Memory is Limited

We report new research results relevant to AI planning in our paper, "Depth-First Memory-Limited AND/OR Search and Unsolvability in Cyclic Search Spaces," presented at the International Joint Conference on Artificial Intelligence, IJCAI-19.

Continue reading