search

facebookresearch / vq2d_cvpr

This repo contains the code for the recipe of the winning entry to the Ego4d VQ2D challenge at CVPR 2022.
MIT License
39 stars 6 forks source link

How to inference a single image with given model? #4

Closed LLsmile closed 1 year ago

frostinassiky commented 1 year ago

Hi @LLsmile, thanks for the quesiton. Although the whole pipeline is designed to query from videos, you can also inference a single image. Please take a look at function perform_retrieval, where you may set clip_frames as a one-element list.

An example of calling this function can be found here.