{"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/fine-tune-the-pretrained-atst-model-for-sound","title":"Fine-tune the pretrained ATST model for sound event detection","arxiv_id":"2309.08153","date":"2023-09-15","proceeding":null,"authors":["Nian Shao","Xian Li","Xiaofei Li"],"abstract":"Sound event detection (SED) often suffers from the data deficiency problem. The recent baseline system in the DCASE2023 challenge task 4 leverages the large pretrained self-supervised learning (SelfSL) models to mitigate such restriction, where the pretrained models help to produce more discriminative features for SED. However, the pretrained models are regarded as a frozen feature extractor in the challenge baseline system and most of the challenge submissions, and fine-tuning of the pretrained models has been rarely studied. In this work, we study the fine-tuning method of the pretrained models for SED. We first introduce ATST-Frame, our newly proposed SelfSL model, to the SED system. ATST-Frame was especially designed for learning frame-level representations of audio signals and obtained state-of-the-art (SOTA) performances on a series of downstream tasks. We then propose a fine-tuning method for ATST-Frame using both (in-domain) unlabelled and labelled SED data. Our experiments show that, the proposed method overcomes the overfitting problem when fine-tuning the large pretrained network, and our SED system obtains new SOTA results of 0.587/0.812 PSDS1/PSDS2 scores on the DCASE challenge task 4 dataset.","url_abs":"https://arxiv.org/abs/2309.08153v2","url_pdf":"https://arxiv.org/pdf/2309.08153v2.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":"fine-tune-the-pretrained-atst-model-for-sound","repo_url":"https://github.com/Audio-WestlakeU/ATST-SED","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"event-detection","task_name":"Event Detection"},{"task_slug":"self-supervised-learning","task_name":"Self-Supervised Learning"},{"task_slug":"sound-event-detection","task_name":"Sound Event Detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/sound-event-detection-on-desed","task":"Sound Event Detection","dataset":"DESED","model":"ATST-SED","rank_in_archive_order":1,"of":13,"metrics":{"PSDS1":"0.583","PSDS2":"0.810","event-based F1 score":"63.4"},"uses_additional_data":true}],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2309.08153","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2309.08153"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Audio-WestlakeU/ATST-SED","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_fixture":1,"ran":9},"by_repo_kind":{"official":{"samples":10,"ran":10,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"55120f2026b56aa2","entry":"drop_path","repo":"Audio-WestlakeU/ATST-SED","repo_kind":"official","path":"desed_task/nnet/atst/transformer.py","file_url":"https://github.com/Audio-WestlakeU/ATST-SED/blob/HEAD/desed_task/nnet/atst/transformer.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"55120f2026b56aa2"}},{"code_sha256_prefix":"2e55511eb5179c63","entry":"get_attention_mask","repo":"Audio-WestlakeU/ATST-SED","repo_kind":"official","path":"desed_task/nnet/atst/transformer.py","file_url":"https://github.com/Audio-WestlakeU/ATST-SED/blob/HEAD/desed_task/nnet/atst/transformer.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"2e55511eb5179c63"}},{"code_sha256_prefix":"38d5542f89a80703","entry":"get_cls_avg","repo":"Audio-WestlakeU/ATST-SED","repo_kind":"official","path":"desed_task/nnet/atst/audio_transformer.py","file_url":"https://github.com/Audio-WestlakeU/ATST-SED/blob/HEAD/desed_task/nnet/atst/audio_transformer.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"38d5542f89a80703"}},{"code_sha256_prefix":"4256112ba8f8c7ab","entry":"get_num_patches","repo":"Audio-WestlakeU/ATST-SED","repo_kind":"official","path":"desed_task/nnet/atst/audio_transformer.py","file_url":"https://github.com/Audio-WestlakeU/ATST-SED/blob/HEAD/desed_task/nnet/atst/audio_transformer.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4256112ba8f8c7ab"}},{"code_sha256_prefix":"4450267615e72051","entry":"mixup","repo":"Audio-WestlakeU/ATST-SED","repo_kind":"official","path":"desed_task/data_augm.py","file_url":"https://github.com/Audio-WestlakeU/ATST-SED/blob/HEAD/desed_task/data_augm.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4450267615e72051"}},{"code_sha256_prefix":"4b00ec1c23e7e9e8","entry":"mixup_w_pretrained","repo":"Audio-WestlakeU/ATST-SED","repo_kind":"official","path":"desed_task/data_augm.py","file_url":"https://github.com/Audio-WestlakeU/ATST-SED/blob/HEAD/desed_task/data_augm.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4b00ec1c23e7e9e8"}},{"code_sha256_prefix":"6d5765c59e00e650","entry":"pad_audio","repo":"Audio-WestlakeU/ATST-SED","repo_kind":"official","path":"desed_task/dataio/datasets_atst_sed.py","file_url":"https://github.com/Audio-WestlakeU/ATST-SED/blob/HEAD/desed_task/dataio/datasets_atst_sed.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6d5765c59e00e650"}},{"code_sha256_prefix":"0cc07361c4eb65c4","entry":"process_labels","repo":"Audio-WestlakeU/ATST-SED","repo_kind":"official","path":"desed_task/dataio/datasets_atst_sed.py","file_url":"https://github.com/Audio-WestlakeU/ATST-SED/blob/HEAD/desed_task/dataio/datasets_atst_sed.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0cc07361c4eb65c4"}},{"code_sha256_prefix":"01e1fd4bd6d27829","entry":"to_mono","repo":"Audio-WestlakeU/ATST-SED","repo_kind":"official","path":"desed_task/dataio/datasets_atst_sed.py","file_url":"https://github.com/Audio-WestlakeU/ATST-SED/blob/HEAD/desed_task/dataio/datasets_atst_sed.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"01e1fd4bd6d27829"}},{"code_sha256_prefix":"7810fed39bf5c241","entry":"trunc_normal_","repo":"Audio-WestlakeU/ATST-SED","repo_kind":"official","path":"desed_task/nnet/atst/audio_transformer.py","file_url":"https://github.com/Audio-WestlakeU/ATST-SED/blob/HEAD/desed_task/nnet/atst/audio_transformer.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"7810fed39bf5c241"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}