STDD improves zero-shot action recognition by adding a parameter-free space-time cross attention to CLIP and augmenting text prompts with an LLM-built action semantic knowledge graph.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Building a Multi-modal Spatiotemporal Expert for Zero-shot Action Recognition with CLIP
STDD improves zero-shot action recognition by adding a parameter-free space-time cross attention to CLIP and augmenting text prompts with an LLM-built action semantic knowledge graph.