Leaner and Faster: Two-Stage Model Compression for Lightweight Text-Image Retrieval
This addresses the practical deployment of text-image retrieval on mobile devices, though it is incremental as it builds on existing pre-trained models.
The paper tackled the problem of high memory and indexing time in dual-encoder text-image retrieval models like CLIP, resulting in a compressed model that is 39% smaller, 1.6x/2.9x faster for image/text processing, and performs on par or better on Flickr30K and MSCOCO benchmarks.
Current text-image approaches (e.g., CLIP) typically adopt dual-encoder architecture using pre-trained vision-language representation. However, these models still pose non-trivial memory requirements and substantial incremental indexing time, which makes them less practical on mobile devices. In this paper, we present an effective two-stage framework to compress large pre-trained dual-encoder for lightweight text-image retrieval. The resulting model is smaller (39% of the original), faster (1.6x/2.9x for processing image/text respectively), yet performs on par with or better than the original full model on Flickr30K and MSCOCO benchmarks. We also open-source an accompanying realistic mobile image search application.