{"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/temporal-knowledge-distillation-for-on-device","title":"Temporal Knowledge Distillation for On-device Audio Classification","arxiv_id":"2110.14131","date":"2021-10-27","proceeding":null,"authors":["Kwanghee Choi","Martin Kersner","Jacob Morton","Buru Chang"],"abstract":"Improving the performance of on-device audio classification models remains a challenge given the computational limits of the mobile environment. Many studies leverage knowledge distillation to boost predictive performance by transferring the knowledge from large models to on-device models. However, most lack a mechanism to distill the essence of the temporal information, which is crucial to audio classification tasks, or similar architecture is often required. In this paper, we propose a new knowledge distillation method designed to incorporate the temporal knowledge embedded in attention weights of large transformer-based models into on-device models. Our distillation method is applicable to various types of architectures, including the non-attention-based architectures such as CNNs or RNNs, while retaining the original network architecture during inference. Through extensive experiments on both an audio event detection dataset and a noisy keyword spotting dataset, we show that our proposed method improves the predictive performance across diverse on-device architectures.","url_abs":"https://arxiv.org/abs/2110.14131v2","url_pdf":"https://arxiv.org/pdf/2110.14131v2.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":[],"tasks":[{"task_slug":"audio-classification","task_name":"Audio Classification"},{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"event-detection","task_name":"Event Detection"},{"task_slug":"keyword-spotting","task_name":"Keyword Spotting"},{"task_slug":"knowledge-distillation","task_name":"Knowledge Distillation"}],"methods":[{"method_slug":"knowledge-distillation","method_name":"Knowledge Distillation"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/audio-classification-on-fsd50k","task":"Audio Classification","dataset":"FSD50K","model":"Temporal Knowledge Distillation for On-device Audio Classification","rank_in_archive_order":8,"of":10,"metrics":{"mAP":"54.8"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2110.14131","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}