Next
Livestream will start soon!
Livestream has already ended.
Presentation has not been recorded yet!
  • title: Non-Monotonic Latent Alignments for CTC-Based NAT
      0:00 / 0:00
      • Report Issue
      • Settings
      • Playlists
      • Bookmarks
      • Subtitles Off
      • Playback rate
      • Quality
      • Settings
      • Debug information
      • Server sl-yoda-v2-stream-002-alpha.b-cdn.net
      • Subtitles size Medium
      • Bookmarks
      • Server
      • sl-yoda-v2-stream-002-alpha.b-cdn.net
      • sl-yoda-v2-stream-002-beta.b-cdn.net
      • 1001562353.rsc.cdn77.org
      • 1075090661.rsc.cdn77.org
      • Subtitles
      • Off
      • English
      • Playback rate
      • Quality
      • Subtitles size
      • Large
      • Medium
      • Small
      • Mode
      • Video Slideshow
      • Audio Slideshow
      • Slideshow
      • Video
      My playlists
        Bookmarks
          00:00:00
            Non-Monotonic Latent Alignments for CTC-Based NAT
            • Settings
            • Sync diff
            • Quality
            • Settings
            • Server
            • Quality
            • Server

            Non-Monotonic Latent Alignments for CTC-Based NAT

            Dec 6, 2022

            Speakers

            CS

            Chenze Shao

            Speaker · 0 followers

            YF

            Yang Feng

            Speaker · 0 followers

            About

            Non-autoregressive translation (NAT) models are typically trained with the cross-entropy loss, which forces the model outputs to be aligned verbatim with the target sentence and will highly penalize small shifts in word positions. Latent alignment models relax the explicit alignment by marginalizing out all monotonic latent alignments with the CTC loss. However, they cannot handle non-monotonic alignments, which is non-negligible as there is typically global word reordering in machine translatio…

            Organizer

            N2
            N2

            NeurIPS 2022

            Account · 952 followers

            Like the format? Trust SlidesLive to capture your next event!

            Professional recording and live streaming, delivered globally.

            Sharing

            Recommended Videos

            Presentations on similar topic, category or speaker

            Program Synthesis for Integer Sequence Generation
            05:10

            Program Synthesis for Integer Sequence Generation

            Natasha Butt, …

            N2
            N2
            NeurIPS 2022 2 years ago

            Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%

            Theseus: A Library for Differentiable Nonlinear Optimization
            04:41

            Theseus: A Library for Differentiable Nonlinear Optimization

            Luis Pineda, …

            N2
            N2
            NeurIPS 2022 2 years ago

            Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%

            “Today we talk to the machine” - Unveiling data for providing micro-credit loans using conversational systems.
            05:32

            “Today we talk to the machine” - Unveiling data for providing micro-credit loans using conversational systems.

            Heloisa Candello, …

            N2
            N2
            NeurIPS 2022 2 years ago

            Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%

            On the Frequency Bias of Coordinate MLPs
            04:55

            On the Frequency Bias of Coordinate MLPs

            Sameera Ramasinghe, …

            N2
            N2
            NeurIPS 2022 2 years ago

            Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%

            Lifelong Learning and Supporting Memory
            35:08

            Lifelong Learning and Supporting Memory

            Hava Siegelmann

            N2
            N2
            NeurIPS 2022 2 years ago

            Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%

            Missing Traffic Data Imputation Using Multi-Trajectory Parameter Transferred LSTM
            03:11

            Missing Traffic Data Imputation Using Multi-Trajectory Parameter Transferred LSTM

            Jungmin Kwon, …

            N2
            N2
            NeurIPS 2022 2 years ago

            Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%

            Interested in talks like this? Follow NeurIPS 2022