Infoseek
PulseAugur coverage of Infoseek — every cluster mentioning Infoseek across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
New 'Ground Then Rank' method boosts knowledge-based visual question answering
Researchers have developed a new framework called "Ground Then Rank" (GTR) to improve Knowledge-Based Visual Question Answering (KB-VQA) performance. This method decouples entity identification from evidence ranking, ad…
-
New MAD-RAG method tackles Attention Distraction in LVLMs
Researchers have identified a new failure mode in retrieval-augmented large vision-language models (LVLMs) called Attention Distraction (AD). This occurs when highly relevant retrieved text globally suppresses visual at…
-
New VQA methods enhance explainability and knowledge integration for multimodal LLMs
Researchers have developed CoExVQA, a new framework for Document Visual Question Answering (DocVQA) that enhances explainability by breaking down the reasoning process. This method first identifies relevant evidence, th…