Pyra: Parallel yielding re-activation for training-inference efficient task adaptation

Y **ong, H Chen, T Hao, Z Lin, J Han, Y Zhang… - … on Computer Vision, 2024 - Springer
Recently, the scale of transformers has grown rapidly, which introduces considerable
challenges in terms of training overhead and inference efficiency in the scope of task …

Parameter efficient fine-tuning via cross block orchestration for segment anything model

Z Peng, Z Xu, Z Zeng, L **e, Q Tian… - Proceedings of the …, 2024 - openaccess.thecvf.com
Parameter-efficient fine-tuning (PEFT) is an effective methodology to unleash the potential of
large foundation models in novel scenarios with limited training data. In the computer vision …

Image compression for machine and human vision with spatial-frequency adaptation

H Li, S Li, S Ding, W Dai, M Cao, C Li, J Zou… - … on Computer Vision, 2024 - Springer
Image compression for machine and human vision (ICMH) has gained increasing attention
in recent years. Existing ICMH methods are limited by high training and storage overheads …

Revisiting the power of prompt for visual tuning

Y Wang, L Cheng, C Fang, D Zhang, M Duan… - arxiv preprint arxiv …, 2024 - arxiv.org
Visual prompt tuning (VPT) is a promising solution incorporating learnable prompt tokens to
customize pre-trained models for downstream tasks. However, VPT and its variants often …

V-PETL Bench: A Unified Visual Parameter-Efficient Transfer Learning Benchmark

Y **n, S Luo, X Liu, Y Du, H Zhou, X Cheng… - The Thirty-eight …, 2024 - openreview.net
Parameter-efficient transfer learning (PETL) methods show promise in adapting a pre-
trained model to various downstream tasks while training only a few parameters. In the …

LSPT: Long-term Spatial Prompt Tuning for Visual Representation Learning

S Mo, Y Wang, X Luo, D Li - arxiv preprint arxiv:2402.17406, 2024 - arxiv.org
Visual Prompt Tuning (VPT) techniques have gained prominence for their capacity to adapt
pre-trained Vision Transformers (ViTs) to downstream visual tasks using specialized …

VioLET: Vision-Language Efficient Tuning with Collaborative Multi-modal Gradients

Y Wang, Y Liu, X Zhang, J Li, B Shi, C Li… - Proceedings of the 31st …, 2023 - dl.acm.org
Parameter-Efficient Tuning (PET) has emerged as a leading advancement in both Natural
Language Processing and Computer Vision, enabling efficient accommodation of …

Learning dual updatable memory modules for video anomaly detection

L Zhang, S Li, Y Cheng, X Luo, X Liu - Multimedia Systems, 2025 - Springer
We propose a novel video anomaly detection method that leverages two updatable memory
modules to learn and update prototypical patterns of normal and abnormal data within an …

iVPT: Improving Task-relevant Information Sharing in Visual Prompt Tuning by Cross-layer Dynamic Connection

N Zhou, J Chen, D Huang - arxiv preprint arxiv:2404.05207, 2024 - arxiv.org
Recent progress has shown great potential of visual prompt tuning (VPT) when adapting pre-
trained vision transformers to various downstream tasks. However, most existing solutions …

BarLeRIa: An Efficient Tuning Framework for Referring Image Segmentation

Y Wang, J Li, X Zhang, B Shi, C Li, W Dai… - The Twelfth International … - openreview.net
Pre-training followed by full fine-tuning has gradually been substituted by Parameter-
Efficient Tuning (PET) in the field of computer vision. PET has gained popularity, especially …