Leveraging Weak Tags for Efficient Image Retrieval via Deep Hyperspherical Quantization
This paper proposes Weakly-Supervised Deep Hyperspherical Quantization (WSDHQ), the first work to address the problem of learning deep quantization from weakly-tagged images without using ground-truth labels.
Revisiting Google Landmarks v2 Dataset: Importance of Avoiding Class Overlap Between Training and Evaluation Sets for Image Retrieval
Removing class overlap between the training set (Google Landmarks v2 clean) and the evaluation sets (Revisited Oxford and Paris) leads to a dramatic drop in performance across state-of-the-art image retrieval methods, highlighting the critical importance of avoiding such overlap. We introduce CiDeR, a single-stage, end-to-end pipeline that detects objects of interest and extracts a global image representation without requiring location supervision, outperforming previous state-of-the-art methods on both the original and the new, revisited dataset.
A Multimodal Approach for Cross-Domain Image Retrieval Study at Imperial College London
Proposing a novel caption-matching method for cross-domain image retrieval using multimodal language-vision architectures.
Fine-Grained Image Retrieval with Sketch and Text Duet
Combining sketches and text for precise image retrieval.
© 2024 by Linnk AI