{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,1]],"date-time":"2026-06-01T23:33:16Z","timestamp":1780356796543,"version":"3.54.1"},"reference-count":20,"publisher":"IEEE","license":[{"start":{"date-parts":[[2022,5,23]],"date-time":"2022-05-23T00:00:00Z","timestamp":1653264000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2022,5,23]],"date-time":"2022-05-23T00:00:00Z","timestamp":1653264000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2022,5,23]]},"DOI":"10.1109\/icassp43922.2022.9747759","type":"proceedings-article","created":{"date-parts":[[2022,4,27]],"date-time":"2022-04-27T15:50:34Z","timestamp":1651074634000},"page":"7157-7161","source":"Crossref","is-referenced-by-count":15,"title":["Tie Your Embeddings Down: Cross-Modal Latent Spaces for End-to-end Spoken Language Understanding"],"prefix":"10.1109","author":[{"given":"Bhuvan","family":"Agrawal","sequence":"first","affiliation":[{"name":"Amazon,Alexa AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Markus","family":"Muller","sequence":"additional","affiliation":[{"name":"Amazon,Alexa AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Samridhi","family":"Choudhary","sequence":"additional","affiliation":[{"name":"Amazon,Alexa AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Martin","family":"Radfar","sequence":"additional","affiliation":[{"name":"Amazon,Alexa AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Athanasios","family":"Mouchtaris","sequence":"additional","affiliation":[{"name":"Amazon,Alexa AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ross","family":"McGowan","sequence":"additional","affiliation":[{"name":"Amazon,Alexa AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nathan","family":"Susanj","sequence":"additional","affiliation":[{"name":"Amazon,Alexa AI"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Siegfried","family":"Kunzmann","sequence":"additional","affiliation":[{"name":"Amazon,Alexa AI"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2158"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2366"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-2929"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP40776.2020.9053281"},{"key":"ref14","article-title":"Bert: Pre-training of deep bidirectional transformers for language understanding","author":"devlin","year":"2018"},{"key":"ref15","article-title":"Spoken language understanding on the edge","author":"saade","year":"2018"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.199"},{"key":"ref17","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-02172-5_2"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2015.7298682"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.541"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2018.8461785"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/SLT.2018.8639043"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2018.8461718"},{"key":"ref5","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-2396"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.1109\/SLT.2018.8639513"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/ASRU.2017.8268987"},{"key":"ref2","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2020-1963"},{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.5626\/JCSE.2010.4.1.001"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2019-1832"},{"key":"ref20","article-title":"Huggingface’s transformers: State-of-the-art natural language processing","author":"wolf","year":"2019"}],"event":{"name":"ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","location":"Singapore, Singapore","start":{"date-parts":[[2022,5,23]]},"end":{"date-parts":[[2022,5,27]]}},"container-title":["ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/9745891\/9746004\/09747759.pdf?arnumber=9747759","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,8,22]],"date-time":"2022-08-22T16:14:27Z","timestamp":1661184867000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/9747759\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,5,23]]},"references-count":20,"URL":"https:\/\/doi.org\/10.1109\/icassp43922.2022.9747759","relation":{},"subject":[],"published":{"date-parts":[[2022,5,23]]}}}