Image Hashing via Cross-View Code Alignment in the Age of Foundation Models

2025-10-31

AI Summary

In the era of foundation models, large-scale image retrieval faces the challenge of learning hash representations that simultaneously achieve compactness and discriminability. To address this, we propose CroVCA—a cross-view code alignment framework—that replaces multi-objective optimization and complex pipelines with a single binary cross-entropy loss jointly regularized by coding-rate maximization, thereby unifying binary code alignment and diversity control. We design a lightweight HashCoder MLP network incorporating batch normalization and LoRA-based fine-tuning, enabling efficient encoder adaptation while freezing backbone features. Evaluated on standard benchmarks, CroVCA achieves state-of-the-art performance within only five training epochs: for 16-bit hashing, it requires less than two minutes for unsupervised hashing on COCO and approximately three minutes for supervised hashing on ImageNet-100—significantly improving both training efficiency and retrieval accuracy.

Technology Category

• Computer Vision: Image and Video Retrieval
• Machine Learning: Representation Learning
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Efficient large-scale retrieval requires representations that are both compact and discriminative. Foundation models provide powerful visual and multimodal embeddings, but nearest neighbor search in these high-dimensional spaces is computationally expensive. Hashing offers an efficient alternative by enabling fast Hamming distance search with binary codes, yet existing approaches often rely on complex pipelines, multi-term objectives, designs specialized for a single learning paradigm, and long training times. We introduce CroVCA (Cross-View Code Alignment), a simple and unified principle for learning binary codes that remain consistent across semantically aligned views. A single binary cross-entropy loss enforces alignment, while coding-rate maximization serves as an anti-collapse regularizer to promote balanced and diverse codes. To implement this, we design HashCoder, a lightweight MLP hashing network with a final batch normalization layer to enforce balanced codes. HashCoder can be used as a probing head on frozen embeddings or to adapt encoders efficiently via LoRA fine-tuning. Across benchmarks, CroVCA achieves state-of-the-art results in just 5 training epochs. At 16 bits, it particularly well-for instance, unsupervised hashing on COCO completes in under 2 minutes and supervised hashing on ImageNet100 in about 3 minutes on a single GPU. These results highlight CroVCA's efficiency, adaptability, and broad applicability.

Problem

• Learning compact binary codes for efficient image retrieval
• Aligning hash codes across semantically consistent image views
• Overcoming computational complexity of high-dimensional foundation model embeddings

Innovation

• Cross-View Code Alignment for binary code consistency
• Single binary cross-entropy loss with coding-rate regularization
• Lightweight HashCoder network enabling rapid frozen or LoRA training