@inproceedings{Singla2020End2EndBehavioralCoding,
 abstract = {Spoken language understanding tasks usually rely on pipelines involving complex processing blocks such as voice activity detection, speaker diarization and Automatic speech recognition (ASR). We propose a novel framework for predicting utterance level labels directly from speech features, thus removing the dependency on first generating transcripts, and transcription free behavioral coding. Our classifier uses a pretrained Speech-2-Vector encoder as bottleneck to generate word-level representations from speech features. This pretrained encoder learns to encode speech features for a word using an objective similar to Word2Vec. Our proposed approach just uses speech features and word segmentation information for predicting spoken utterance-level target labels. We show that our model achieves competitive results to other state-of-the-art approaches which use transcribed text for the task of predicting psychotherapy-relevant behavior codes.},
 address = {Online},
 author = {Singla, Karan and Chen, Zhuohao and Atkins, David and Narayanan, Shrikanth},
 booktitle = {Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics},
 doi = {10.18653/v1/2020.acl-main.351},
 link = {https://www.aclweb.org/anthology/2020.acl-main.351.pdf},
 month = {Jul},
 pages = {3797-3803},
 publisher = {Association for Computational Linguistics},
 title = {Towards end-2-end learning for predicting behavior codes from spoken utterances in psychotherapy conversations},
 year = {2020}
}

{"_id":"wNLy9NQQt7PatTnPf","bibbaseid":"singla-chen-atkins-narayanan-towardsend2endlearningforpredictingbehaviorcodesfromspokenutterancesinpsychotherapyconversations-2020","author_short":["Singla, K.","Chen, Z.","Atkins, D.","Narayanan, S."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","abstract":"Spoken language understanding tasks usually rely on pipelines involving complex processing blocks such as voice activity detection, speaker diarization and Automatic speech recognition (ASR). We propose a novel framework for predicting utterance level labels directly from speech features, thus removing the dependency on first generating transcripts, and transcription free behavioral coding. Our classifier uses a pretrained Speech-2-Vector encoder as bottleneck to generate word-level representations from speech features. This pretrained encoder learns to encode speech features for a word using an objective similar to Word2Vec. Our proposed approach just uses speech features and word segmentation information for predicting spoken utterance-level target labels. We show that our model achieves competitive results to other state-of-the-art approaches which use transcribed text for the task of predicting psychotherapy-relevant behavior codes.","address":"Online","author":[{"propositions":[],"lastnames":["Singla"],"firstnames":["Karan"],"suffixes":[]},{"propositions":[],"lastnames":["Chen"],"firstnames":["Zhuohao"],"suffixes":[]},{"propositions":[],"lastnames":["Atkins"],"firstnames":["David"],"suffixes":[]},{"propositions":[],"lastnames":["Narayanan"],"firstnames":["Shrikanth"],"suffixes":[]}],"booktitle":"Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics","doi":"10.18653/v1/2020.acl-main.351","link":"https://www.aclweb.org/anthology/2020.acl-main.351.pdf","month":"Jul","pages":"3797-3803","publisher":"Association for Computational Linguistics","title":"Towards end-2-end learning for predicting behavior codes from spoken utterances in psychotherapy conversations","year":"2020","bibtex":"@inproceedings{Singla2020End2EndBehavioralCoding,\n abstract = {Spoken language understanding tasks usually rely on pipelines involving complex processing blocks such as voice activity detection, speaker diarization and Automatic speech recognition (ASR). We propose a novel framework for predicting utterance level labels directly from speech features, thus removing the dependency on first generating transcripts, and transcription free behavioral coding. Our classifier uses a pretrained Speech-2-Vector encoder as bottleneck to generate word-level representations from speech features. This pretrained encoder learns to encode speech features for a word using an objective similar to Word2Vec. Our proposed approach just uses speech features and word segmentation information for predicting spoken utterance-level target labels. We show that our model achieves competitive results to other state-of-the-art approaches which use transcribed text for the task of predicting psychotherapy-relevant behavior codes.},\n address = {Online},\n author = {Singla, Karan and Chen, Zhuohao and Atkins, David and Narayanan, Shrikanth},\n booktitle = {Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics},\n doi = {10.18653/v1/2020.acl-main.351},\n link = {https://www.aclweb.org/anthology/2020.acl-main.351.pdf},\n month = {Jul},\n pages = {3797-3803},\n publisher = {Association for Computational Linguistics},\n title = {Towards end-2-end learning for predicting behavior codes from spoken utterances in psychotherapy conversations},\n year = {2020}\n}\n\n","author_short":["Singla, K.","Chen, Z.","Atkins, D.","Narayanan, S."],"bibbaseid":"singla-chen-atkins-narayanan-towardsend2endlearningforpredictingbehaviorcodesfromspokenutterancesinpsychotherapyconversations-2020","role":"author","urls":{},"metadata":{"authorlinks":{}}},"bibtype":"inproceedings","biburl":"https://bibbase.org/f/nWhKb4SffvhfreEmj/shri-isi-edu.bib","dataSources":["P3nQrSLkFzGGSmKJQ","Reikhy6EiDXFTcuR9"],"keywords":[],"search_terms":["towards","end","end","learning","predicting","behavior","codes","spoken","utterances","psychotherapy","conversations","singla","chen","atkins","narayanan"],"title":"Towards end-2-end learning for predicting behavior codes from spoken utterances in psychotherapy conversations","year":2020}