Stéphane Lathuilière, Elisa Ricci, Subhankar Roy, Alessandro Conti, Paolo Rota, Giacomo Zara
We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. If it links a repository, it is listed below.
Source-Free Video Unsupervised Domain Adaptation (SFVUDA) task consists in adapting an action recognition model, trained on a labelled source dataset, to an unlabelled target dataset, without accessing the actual source data. The previous approaches have attempted to address SFVUDA by leveraging self-supervision (e.g., enforcing temporal consistency) derived from the target data itself. In this work, we take an orthogonal approach by exploiting "web-supervision" from Large Language-Vision Models (LLVMs), driven by the rationale that LLVMs contain a rich world prior surprisingly robust to domainshift. We showcase the unreasonable effectiveness of integrating LLVMs for SFVUDA by devising an intuitive and parameter-efficient method, which we name Domain Adaptation with Large Language-Vision models (DALL-V), that distills the world prior and complementary source model information into a student network tailored for the target. Despite the simplicity, DALL-V 1 achieves significant improvement over state-of-the-art SFVUDA methods.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2308.09139")
get_code_for_paper("2308.09139")
have("2308.09139")
Connect an agent — have() is free.