{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/self-paced-learning-with-adaptive-deep-visual","title":"Self-Paced Learning with Adaptive Deep Visual Embeddings","arxiv_id":"1807.09200","date":"2018-07-24","proceeding":null,"authors":["Vithursan Thangarasa","Graham W. Taylor"],"abstract":"Selecting the most appropriate data examples to present a deep neural network\n(DNN) at different stages of training is an unsolved challenge. Though\npractitioners typically ignore this problem, a non-trivial data scheduling\nmethod may result in a significant improvement in both convergence and\ngeneralization performance. In this paper, we introduce Self-Paced Learning\nwith Adaptive Deep Visual Embeddings (SPL-ADVisE), a novel end-to-end training\nprotocol that unites self-paced learning (SPL) and deep metric learning (DML).\nWe leverage the Magnet Loss to train an embedding convolutional neural network\n(CNN) to learn a salient representation space. The student CNN classifier\ndynamically selects similar instance-level training examples to form a\nmini-batch, where the easiness from the cross-entropy loss and the true\ndiverseness of examples from the learned metric space serve as sample\nimportance priors. To demonstrate the effectiveness of SPL-ADVisE, we use deep\nCNN architectures for the task of supervised image classification on several\ncoarse- and fine-grained visual recognition datasets. Results show that, across\nall datasets, the proposed method converges faster and reaches a higher final\naccuracy than other SPL variants, particularly on fine-grained classes.","url_abs":"http://arxiv.org/abs/1807.09200v1","url_pdf":"http://arxiv.org/pdf/1807.09200v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"self-paced-learning-with-adaptive-deep-visual","repo_url":"https://github.com/vithursant/SPL-ADVisE","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"fine-grained-visual-recognition","task_name":"Fine-Grained Visual Recognition"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"metric-learning","task_name":"Metric Learning"},{"task_slug":"scheduling","task_name":"Scheduling"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1807.09200","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}