On a curated 20-query subset of Market1501, PersonViT strongly outperforms ChatGPT-4o, Gemini-2.0-Flash, Claude 3.5 Sonnet, and Qwen-VL-Max in separating genuine from impostor person matches.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Human Re-ID Meets LVLMs: What can we expect?
On a curated 20-query subset of Market1501, PersonViT strongly outperforms ChatGPT-4o, Gemini-2.0-Flash, Claude 3.5 Sonnet, and Qwen-VL-Max in separating genuine from impostor person matches.