computer-vision-5
Video-Adverb Retrieval (Unseen Compositions)
3260 papers • 126 benchmarks • 313 datasets
The task aims to recognize adverbs beyond seen adverb-action compositions, i.e. compositions that were not seen during training.
(Image credit: Papersgraph)
Benchmarks
These leaderboards are used to track progress in video-adverb-retrieval-unseen-compositions-5
Trend
Dataset
Best Model
Actions
VATEX Adverbs
VATEX Adverbs
MSR-VTT Adverbs
MSR-VTT Adverbs
ActivityNet Adverbs
Libraries
iUse these libraries to find video-adverb-retrieval-unseen-compositions-5 models and implementations
Datasets
Subtasks
No subtasks available.
Most implemented papers
Video-adverb retrieval with compositional adverb-action embeddings
Zeynep Akata, A. S. Koepke, Thomas Hummel, Otniel-Bogdan Mercea•Mon Sep 25 2023
This work proposes a framework for video-to-adverb retrieval that aligns video embeddings with their matching compositional adverb-action text embedding in a joint embedding space and outperforms all prior works for the generalisation task of retrieving adverbs from videos for unseen adverbs-action compositions.
1