Pith. sign in

REVIEW 2 cited by

Test-Time Adaptation with CLIP Reward for Zero-Shot Generalization in Vision-Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2305.18010 v2 pith:JH6WLZO3 submitted 2023-05-29 cs.CV cs.MM

Test-Time Adaptation with CLIP Reward for Zero-Shot Generalization in Vision-Language Models

classification cs.CV cs.MM
keywords modelgeneralizationrlcfzero-shotcliprewardvlmsfeedback
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

One fascinating aspect of pre-trained vision-language models~(VLMs) learning under language supervision is their impressive zero-shot generalization capability. However, this ability is hindered by distribution shifts between the training and testing data. Previous test time adaptation~(TTA) methods for VLMs in zero-shot classification rely on minimizing the entropy of model outputs, tending to be stuck in incorrect model predictions. In this work, we propose TTA with feedback to rectify the model output and prevent the model from becoming blindly confident. Specifically, a CLIP model is adopted as the reward model during TTA and provides feedback for the VLM. Given a single test sample, the VLM is forced to maximize the CLIP reward between the input and sampled results from the VLM output distribution. The proposed \textit{reinforcement learning with CLIP feedback~(RLCF)} framework is highly flexible and universal. Beyond the classification task, with task-specific sampling strategies and a proper reward baseline choice, RLCF can be easily extended to not only discrimination tasks like retrieval but also generalization tasks like image captioning, improving the zero-shot generalization capacity of VLMs. According to the characteristics of these VL tasks, we build different fully TTA pipelines with RLCF to improve the zero-shot generalization ability of various VLMs. Extensive experiments along with promising empirical results demonstrate the effectiveness of RLCF. The code is available at https://github.com/mzhaoshuai/RLCF.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Turning Adaptation into Assets: Cross-Domain Bridging for Online Vision-Language Navigation

    cs.RO 2026-05 unverdicted novelty 6.0

    IDEA is a TTA framework for VLN that builds a dynamic asset library from Fisher-weighted soft prompts and domain coordinates, then uses convex-hull projection for cross-domain bridging and training-free adaptation.

  2. Seeking Consensus: Geometric-Semantic On-the-Fly Recalibration for Open-Vocabulary Remote Sensing Semantic Segmentation

    cs.CV 2026-04 unverdicted novelty 6.0

    SeeCo is a training-free on-the-fly recalibration method using multi-view geometric consistency and adaptive textual calibration to improve open-vocabulary semantic segmentation in remote sensing images.