We present F-VLM, a simple open-vocabulary object detection method built upon Frozen Vision and Language Models. F-VLM simplifies the current multi-stage training pipeline by eliminating the need for knowledge distillation or detection-tailored pretraining. Surprisingly, we observe that a frozen VLM: 1) retains the locality-sensitive features necessary for detection, and 2) is a strong region classifier. We finetune only the detector head and combine the detector and VLM outputs for each region at inference time. F-VLM shows compelling scaling behavior and achieves +6.5 mask AP improvement over the previous state of the art on novel categories of LVIS open-vocabulary detection benchmark. In addition, we demonstrate very competitive results on COCO open-vocabulary detection benchmark and cross-dataset transfer detection, in addition to significant training speed-up and compute savings. Code will be released at the https://sites.google.com/view/f-vlm/home
@article{arxiv.2209.15639,
title = {F-VLM: Open-Vocabulary Object Detection upon Frozen Vision and Language Models},
author = {Weicheng Kuo and Yin Cui and Xiuye Gu and AJ Piergiovanni and Anelia Angelova},
journal= {arXiv preprint arXiv:2209.15639},
year = {2023}
}
Comments
Accepted to ICLR 2023 (https://iclr.cc/Conferences/2023). 20 pages, 7 figures