AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

MobileViCLIP: An Efficient Video-Text Model for Mobile Devices

Min Yang, Zihan Jia, Zhilin Dai, Sheng Guo, Limin Wang

ICCV 2025 · Vision-language and multimodal › CLIP and zero-shot · approach: Fine-grained improvements to CLIP · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add spatiotemporal RepMixer to MobileCLIP (ablation)MobileCLIP · MSR-VTT fine-tuned T2VR@1: 38.4 -> 39.5Improved
add spatiotemporal attention without TPE (ablation)MobileCLIP · MSR-VTT fine-tuned T2VR@1: 38.4 -> 39.1Improved
add learnable temporal positional encodings to spatiotemporal attention (ablation)MobileViCLIP · MSR-VTT fine-tuned T2VR@1: 39.1 -> 39.6Improved
combine RepMixer and spatiotemporal attention with TPE (ablation)MobileViCLIP · MSR-VTT fine-tuned T2VR@1: 39.5 -> 40.1Improved
also train the text branch rather than freezing it (ablation)MobileViCLIP · MSR-VTT fine-tuned T2VR@1: 40.1 -> 39.7Hurt
fine-tune MobileCLIP-S0 on InternVid-10M-FLT (no temporal modules)MobileCLIP-S0 · MSR-VTT zero-shotT2V R@1: 33.2 -> 37.2Improved
add temporal modules on top of InternVid fine-tuningMobileCLIP-S0 · MSR-VTT zero-shotT2V R@1: 37.2 -> 38.7Improved
add temporal modules on top of InternVid fine-tuningMobileCLIP-S2 · MSR-VTT zero-shotT2V R@1: 41.2 -> 42.5Improved
add temporal modules on top of InternVid fine-tuningMobileCLIP-S2 · SSV2-MC zero-shottop-1 accuracy: 44.1 -> 46.1Improved
MobileViCLIP-Small vs InternVideo2-S14 on ActivityNet-style long videoMobileViCLIP-Small · SSV2-MC zero-shottop-1 accuracy: 46.4 -> 46.1Hurt
use MobileViCLIP-Small features instead of MobileCLIP-S0(10M) for temporal groundingCG-DETR · QVHighlightsmAP avg: 43.8 -> 46.3Improved
use MobileViCLIP-Small features instead of CoCa ViT-L/14 for zero-shot temporal action detectionT3AL · THUMOS14avg mAP: 4.8 -> 5.3Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy