Skip to yearly menu bar Skip to main content


Spotlight
in
Workshop: UniReps: Unifying Representations in Neural Models

CLIP meets Model Zoo Experts: Pseudo-Supervision for Visual Enhancement

Mohammadreza (Reza) Salehi · Mehrdad Farajtabar · Maxwell Horton · Fartash Faghri · Hadi Pouransari · Raviteja Vemulapalli · Oncel Tuzel · Ali Farhadi · Mohammad Rastegari · Sachin Mehta


Abstract:

Contrastive language image pretraining (CLIP) is a standard method for training vision-language models. While CLIP is scalable, promptable, and robust to distribution shifts on image classification tasks, it lacks object localization capabilities. This paper studies the following question: Can we augment CLIP training with task-specific vision models from model zoos to improve its visual representations? Towards this end, we leverage open-source task-specific vision models to generate pseudo-labels for an uncurated web-scale image-text dataset. Subsequently, we train CLIP models on these pseudo-labels in addition to the contrastive training on image and text pairs. This simple setup shows substantial improvements of up to 16.3\% across different vision tasks, including segmentation, detection, depth estimation, and surface normal estimation. Importantly, these enhancements are achieved without compromising CLIP's existing capabilities, including its proficiency in promptable zero-shot classification.

Chat is not available.