{"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/titanet-neural-model-for-speaker","title":"TitaNet: Neural Model for speaker representation with 1D Depth-wise separable convolutions and global context","arxiv_id":"2110.04410","date":"2021-10-08","proceeding":null,"authors":["Nithin Rao Koluguri","Taejin Park","Boris Ginsburg"],"abstract":"In this paper, we propose TitaNet, a novel neural network architecture for extracting speaker representations. We employ 1D depth-wise separable convolutions with Squeeze-and-Excitation (SE) layers with global context followed by channel attention based statistics pooling layer to map variable-length utterances to a fixed-length embedding (t-vector). TitaNet is a scalable architecture and achieves state-of-the-art performance on speaker verification task with an equal error rate (EER) of 0.68% on the VoxCeleb1 trial file and also on speaker diarization tasks with diarization error rate (DER) of 1.73% on AMI-MixHeadset, 1.99% on AMI-Lapel and 1.11% on CH109. Furthermore, we investigate various sizes of TitaNet and present a light TitaNet-S model with only 6M parameters that achieve near state-of-the-art results in diarization tasks.","url_abs":"https://arxiv.org/abs/2110.04410v1","url_pdf":"https://arxiv.org/pdf/2110.04410v1.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":"titanet-neural-model-for-speaker","repo_url":"https://github.com/NVIDIA/NeMo","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"titanet-neural-model-for-speaker","repo_url":"https://github.com/Wadaboa/titanet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"speaker-diarization","task_name":"Speaker Diarization"},{"task_slug":"speaker-verification","task_name":"Speaker Verification"},{"task_slug":"speaker-diarization","task_name":"speaker-diarization"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/speaker-diarization-on-ami-lapel","task":"Speaker Diarization","dataset":"AMI Lapel","model":"TitaNet-M (NME-SC)","rank_in_archive_order":1,"of":4,"metrics":{"DER(%)":"1.99"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-ami-lapel","task":"Speaker Diarization","dataset":"AMI Lapel","model":"TitaNet-S (NME-SC)","rank_in_archive_order":2,"of":4,"metrics":{"DER(%)":"2.00"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-ami-lapel","task":"Speaker Diarization","dataset":"AMI Lapel","model":"TitaNet-L (NME-SC)","rank_in_archive_order":3,"of":4,"metrics":{"DER(%)":"2.03"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-ami-lapel","task":"Speaker Diarization","dataset":"AMI Lapel","model":"ECAPA (SC)","rank_in_archive_order":4,"of":4,"metrics":{"DER(%)":"2.36"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-ami-mixheadset","task":"Speaker Diarization","dataset":"AMI MixHeadset","model":"TitaNet-L (NME-SC)","rank_in_archive_order":1,"of":4,"metrics":{"DER(%)":"1.73"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-ami-mixheadset","task":"Speaker Diarization","dataset":"AMI MixHeadset","model":"ECAPA (SC)","rank_in_archive_order":2,"of":4,"metrics":{"DER(%)":"1.78"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-ami-mixheadset","task":"Speaker Diarization","dataset":"AMI MixHeadset","model":"TitaNet-M (NME-SC)","rank_in_archive_order":3,"of":4,"metrics":{"DER(%)":"1.79"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-ami-mixheadset","task":"Speaker Diarization","dataset":"AMI MixHeadset","model":"TitaNet-S (NME-SC)","rank_in_archive_order":4,"of":4,"metrics":{"DER(%)":"2.22"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-callhome-109","task":"Speaker Diarization","dataset":"CALLHOME-109","model":"titanet-s","rank_in_archive_order":1,"of":2,"metrics":{"DER(%)":"1.11"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-ch109","task":"Speaker Diarization","dataset":"CH109","model":"TitaNet-S (NME-SC)","rank_in_archive_order":1,"of":4,"metrics":{"DER(%)":"1.11"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-ch109","task":"Speaker Diarization","dataset":"CH109","model":"TitaNet-M (NME-SC)","rank_in_archive_order":2,"of":4,"metrics":{"DER(%)":"1.13"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-ch109","task":"Speaker Diarization","dataset":"CH109","model":"TitaNet-L (NME-SC)","rank_in_archive_order":3,"of":4,"metrics":{"DER(%)":"1.19"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-ch109","task":"Speaker Diarization","dataset":"CH109","model":"x-vector (PLDA + AHC)","rank_in_archive_order":4,"of":4,"metrics":{"DER(%)":"9.72"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-nist-sre-2000","task":"Speaker Diarization","dataset":"NIST-SRE 2000","model":"x-vector (MCGAN)","rank_in_archive_order":1,"of":5,"metrics":{"DER(%)":"5.73"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-nist-sre-2000","task":"Speaker Diarization","dataset":"NIST-SRE 2000","model":"TitaNet-S (NME-SC)","rank_in_archive_order":2,"of":5,"metrics":{"DER(%)":"6.37"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-nist-sre-2000","task":"Speaker Diarization","dataset":"NIST-SRE 2000","model":"TitaNet-M (NME-SC)","rank_in_archive_order":3,"of":5,"metrics":{"DER(%)":"6.47"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-nist-sre-2000","task":"Speaker Diarization","dataset":"NIST-SRE 2000","model":"TitaNet-L (NME-SC)","rank_in_archive_order":4,"of":5,"metrics":{"DER(%)":"6.73"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-diarization-on-nist-sre-2000","task":"Speaker Diarization","dataset":"NIST-SRE 2000","model":"x-vector (PLDA + AHC)","rank_in_archive_order":5,"of":5,"metrics":{"DER(%)":"8.39"},"uses_additional_data":false},{"leaderboard":"/sota/speaker-verification-on-voxceleb","task":"Speaker Verification","dataset":"VoxCeleb","model":"TitanNet -L","rank_in_archive_order":13,"of":21,"metrics":{"EER":"0.68"},"uses_additional_data":true},{"leaderboard":"/sota/speaker-verification-on-voxceleb","task":"Speaker Verification","dataset":"VoxCeleb","model":"TitanNet -M","rank_in_archive_order":15,"of":21,"metrics":{"EER":"0.81"},"uses_additional_data":true},{"leaderboard":"/sota/speaker-verification-on-voxceleb","task":"Speaker Verification","dataset":"VoxCeleb","model":"TitanNet -S","rank_in_archive_order":19,"of":21,"metrics":{"EER":"1.15"},"uses_additional_data":true}],"syntology":{"syntology_url":"https://syntology.ai/paper/2110.04410","atlas_url":"https://app.syntology.ai/?focus=2110.04410","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2110.04410"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+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/NVIDIA/NeMo","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Wadaboa/titanet","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":4},"by_repo_kind":{"listed":{"samples":4,"ran":4,"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":"c8322253612f1f9f","entry":"collate_fn","repo":"Wadaboa/titanet","repo_kind":"listed","path":"src/datasets.py","file_url":"https://github.com/Wadaboa/titanet/blob/HEAD/src/datasets.py","link_basis":"harvester_set","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":"c8322253612f1f9f"}},{"code_sha256_prefix":"df2af7d3bc841ca8","entry":"copy_example","repo":"Wadaboa/titanet","repo_kind":"listed","path":"src/transforms.py","file_url":"https://github.com/Wadaboa/titanet/blob/HEAD/src/transforms.py","link_basis":"harvester_set","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":"df2af7d3bc841ca8"}},{"code_sha256_prefix":"7d53367371a2e0a3","entry":"get_random_filename","repo":"Wadaboa/titanet","repo_kind":"listed","path":"src/utils.py","file_url":"https://github.com/Wadaboa/titanet/blob/HEAD/src/utils.py","link_basis":"harvester_set","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":"7d53367371a2e0a3"}},{"code_sha256_prefix":"66330d0b22b2081f","entry":"to_numpy","repo":"Wadaboa/titanet","repo_kind":"listed","path":"src/utils.py","file_url":"https://github.com/Wadaboa/titanet/blob/HEAD/src/utils.py","link_basis":"harvester_set","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":"66330d0b22b2081f"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}