Find your Needle: Small Object Image Retrieval via Multi-Object Attention Optimization

We address the challenge of Small Object Image Retrieval (SoIR), where the goal is to retrieve images containing a specific small object, in a cluttered scene. The key challenge in this setting is constructing a single image descriptor, for scalable and efficient search, that effectively represents all objects in the image. In this paper, we first analyze the limitations of existing methods on this challenging task and then introduce new benchmarks to support SoIR evaluation. Next, we introduce Multi-object Attention Optimization (MaO), a novel retrieval framework which incorporates a dedicated multi-object pre-training phase. This is followed by a refinement process that leverages attention-based feature extraction with object masks, integrating them into a single unified image descriptor. Our MaO approach significantly outperforms existing retrieval methods and strong baselines, achieving notable improvements in both zero-shot and lightweight multi-object fine-tuning. We hope this work will pave the way and inspire further research to enhance retrieval performance for this highly practical task. Project page: https://pihash2k.github.io/findyourneedle.github.io. MaO Top-1 Result (for query A) Target is here Query A Query B Top-1 Result (for query B) Target is here MaO Retrieve Retrieve Figure 1: It is highly challenging to find an object instance that appears small in a cluttered scene image, moreover when shuffled in a large corpus of images. Top-1 retrieval results by our Multi-Object Attention Optimization (MaO) method are shown, attained with a single representation per-image, allowing a scalable search.