Salesforce Research at ICML 2022

Conference Overview

This weekend will kick off the thirty-ninth International Conference on Machine Learning (ICML). This conference specifically aims to bring together professionals who are dedicated to the advancement of Machine Learning (ML) in Artificial Intelligence. Participants at ICML come from many different backgrounds, including academic and industrial researchers, entrepreneurs and engineers, and graduate students and postdocs. (ICML Website)

This year, ICML 2022 will take place in a hybrid format, hosting people both virtually and in-person in Baltimore, Maryland from July 17th-23rd, 2022.

Salesforce AI Research Publications at NAACL 2022

Salesforce Research is pleased to announce a total of 2 accepted papers from our team of leading researchers.

Our accepted authors will present their work at ICML through pre-recorded talks and in-person poster sessions during the main conference. We look forward to sharing some of our exciting new research, whether virtually or face-to-face in Baltimore!

Near-Optimal Learning of Extensive-Form Games with Imperfect Information

Yu Bai, Chi Jin, Song Mei, Tiancheng Yu

Ranging from strategic games such as Poker to real-world scenarios such as diplomacy, multi-agent games with imperfect information constitute a central challenge for modern AI. This paper develops new algorithms for solving sequential imperfect-information games with state-of-the-art sample efficiency. Our algorithms yield strong game AI both in online playing and from offline self-training, and are applicable in either competitive or collaborative games.

BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation

Junnan Li, Dongxu Li, Caiming Xiong, Steven Hoi

BLIP is a multimodal AI system with state-of-the-art performance on a wide range of vision-and-language tasks. It proposes a novel method that can effectively learn from large-scale freely-available image-text pairs on the web. BLIP has gained much popularity in the AI research community with 700+ github stars (https://github.com/salesforce/BLIP). BLIP not only advances the research frontier, but also provides a powerful backbone engine for multimodal applications on Salesforce platforms.