This is the official implementation of LLaVA-ReID: Selective Multi-image Questioner for Interactive Person Re-Identification (ICML 2025). tree -L 3 . ├── CUHK-PEDES │ ├── caption_all.json │ ├── imgs ...
Note that the data follows the format for LLaVA-style training. We use slurm_hp_with_mover.sh that runs the code in SLURM with hyper-parameter search. Set the variables correctly for the mode and run ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback