YOLO model parameter efficient fine tuning method based on low-rank self-adaption

By introducing a low-rank adaptive layer into the YOLO model, which only updates the low-rank matrices B and A, and optimizing the multi-scale feature fusion and self-attention mechanism, the problem of balancing parameter efficiency and detection performance in complex scenes of the YOLO model is solved. This achieves efficient fine-tuning and fast adaptation, improving detection accuracy and adaptability.

CN120912899AInactive Publication Date: 2025-11-07BEIJING INSTITUTE OF GRAPHIC COMMUNICATION
0 Cites 2 Cited by

Patent Information

Application Number
CN202510925013.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-04
Publication Date
2025-11-07
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Existing YOLO object detection models struggle to balance parameter efficiency and detection performance, especially in complex scenarios where they lack adaptability. Traditional full-parameter fine-tuning methods are computationally expensive and difficult to deploy and apply in real time on edge devices.

Method used

We employ a low-rank adaptive YOLO model parameter fine-tuning method. By embedding a low-rank adaptive layer into the convolutional layer of the YOLO model, we utilize low-rank matrix factorization to generate trainable weight increment matrices, update only the low-rank matrices B and A, reduce the number of parameters, and optimize multi-scale feature fusion and self-attention mechanisms.

Benefits of technology

It significantly reduces computing resource consumption, improves target detection accuracy and feature capture capabilities in complex scenarios, supports rapid adaptation to downstream tasks, reduces industrial deployment and maintenance costs, and balances real-time detection performance with robustness.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a YOLO model parameter efficient fine tuning method based on low-rank self-adaption. Aiming at the problems that a traditional full-parameter fine tuning method is high in calculation cost and large in storage pressure and an existing efficient parameter fine tuning method is insufficient in adaptation in a visual task, a low-rank adaptive layer is introduced into a convolutional layer of a YOLO model, a weight update quantity is decomposed into a low-rank matrix product, and a feature fusion process is optimized in combination with a lightweight adapter. In specific implementation, customized improvement is performed on a core module of the YOLO model, including multi-scale feature fusion optimization, attention weight dynamic adjustment and the like, and meanwhile, pre-training parameters are frozen and only a low-rank matrix is updated during training. The parameter efficiency and the detection performance are effectively balanced under the condition that only a small number of parameters are finely adjusted, and the method is suitable for edge device deployment and real-time detection scenes.
Need to check novelty before this filing date? Find Prior Art