Making the Most of Text Semantics to Improve Biomedical Vision-Language Processing

Channel:
Microsoft Research
Subscribers:
343,000
Published on ● Video Link: https://www.youtube.com/watch?v=by-JzcUJQpw


Duration: 5:01
845 views
28

Presented by the Microsoft Health Futures team at ECCV 2022 - Poster session #500subs

See more at https://www.microsoft.com/en-us/research/publication/making-the-most-of-text-semantics-to-improve-biomedical-vision-language-processing/



Other Videos By Microsoft Research


2022-10-27Plenary: The fifth paradigm of scientific discovery
2022-10-27Plenary: The future of human health [ASL version]
2022-10-27Plenary: The future of human health
2022-10-19Project Silica 2022
2022-10-18Project Silica Library 2022
2022-10-14A Random Walk from Complexity Theory to ML. With Dr. Neeraj Kayal and Dr. Ravishankar Krishnaswamy
2022-10-14HAMS- Using Smartphones to Make Roads Safer. With Dr. Venkat Padmanabhan and Dr. Akshay Nambi
2022-10-13Microsoft Research 2022 Global PhD Fellowship Recipients
2022-10-10SIKE Channels – zero value side-channel attacks on SIKE
2022-10-06Irrational Design: Considering Reflexivity, Speculation, and Long-term Engagement in HCI
2022-10-04Making the Most of Text Semantics to Improve Biomedical Vision-Language Processing
2022-09-23AI4Science in Conversation: Chris Bishop and Frank Noé
2022-09-21Women in Data Science Fireside Chat with Cathy Tanimura, Eren Manavoglu and Yan Guo
2022-09-20A Prediction Model for Malaria using an Ensemble of Machine Learning & Hydrological Drought Indices
2022-09-13Project Alexandria: Using AI to build a knowledge base fit for humans
2022-09-06Sumit Gulwani: 3-min video introduction
2022-08-09InAs-Al Hybrid Devices Passing the Topological Gap Protocol
2022-07-29“Freedom Dreams”: Imagining Inclusive Technology Futures through Co-Design with Black Americans
2022-07-26Accelerating the Delfs-Galbraith algorithm with fast subfield root detection
2022-07-22At the Frontiers of Science: Exciting Advances in Protein Design
2022-07-223D Face Reconstruction with Dense Landmarks