Dual-tuning: Joint Prototype Transfer And Structure Regularization For Compatible Feature Learning

Abstract

Visual retrieval system faces frequent model update and deployment. It is a heavy workload to re-extract features of the whole database every time.Feature compatibility enables the learned new visual features to be directly compared with the old features stored in the database. In this way, when updating the deployed model, we can bypass the inflexible and time-consuming feature re-extraction process. However, the old feature space that needs to be compatible is not ideal and faces the distribution discrepancy problem with the new space caused by different supervision losses. In this work, we propose a global optimization Dual-Tuning method to obtain feature compatibility against different networks and losses. A feature-level prototype loss is proposed to explicitly align two types of embedding features, by transferring global prototype information. Furthermore, we design a component-level mutual structural regularization to implicitly optimize the feature intrinsic structure. Experime

Stats

• citations3
• S2 citations—
• github stars5
• HF likes0
• heat score6.07
• arxiv keybai2021dual

Related papers

• Towards Backward-compatible Representation Learning (2020)12.93
• Backward-compatible Aligned Representations Via An Orthogonal Transformation Layer (2024)0.00
• Stationary Representations: Optimally Approximating Compatibility And Implications For Improved Model Replacements (2024)5.87
• Forward Compatible Training For Large-scale Embedding Retrieval Systems (2021)8.09
• Reuse Your Features: Unifying Retrieval And Feature-metric Alignment (2022)1.69
• Prototype Perturbation For Relaxing Alignment Constraints In Backward-compatible Learning (2025)0.00
• Learning Compatible Embeddings (2021)13.80
• Unified Representation Learning For Cross Model Compatibility (2020)5.24