Sound Capture and Speech Enhancement for Communication and Distant Speech Recognition

Channel:
Microsoft Research
Subscribers:
344,000
Published on ● Video Link: https://www.youtube.com/watch?v=VGu13TNhezo


Duration: 1:37:59
1,882 views
44

In this talk we will discuss the general architecture of speech enhancement pipelines for the needs of hands-free telecommunication and distant speech recognition. The talk will discuss both classical approaches using statistical signal processing and deep learning using neural networks. It will be illustrated with real-life examples from the speech enhancement audio pipelines in Kinect, HoloLens, and Teams.

PRESENTERS: Dr. Ivan J. Tashev and Dr. Sebastian Braun from Audio and Acoustics Research Group in Microsoft Research – Redmond, WA, USA.



Other Videos By Microsoft Research


2021-05-20Failures of imagination: Discovering and measuring harms in language technologies
2021-05-13Cities Unlocked – Introducing 3D Sound for Greater Mobility and Independence
2021-05-13The Journey to Microsoft Soundscape
2021-05-13Microsoft Soundscape - Lighting up the World with Sound
2021-05-12Platform for Situated Intelligence Workshop | Day 1
2021-05-12Platform for Situated Intelligence Workshop | Day 2
2021-05-03Knowledge Distillation as Semiparametric Inference
2021-05-03Better design, implementation, and testing of async systems with Coyote
2021-05-03Research @Microsoft Research India: interdisciplinary and impactful with Dr. Sriram Rajamani
2021-04-29Virtual Lake Nona Impact Forum “Health Innovation in the New Reality”
2021-04-28Sound Capture and Speech Enhancement for Communication and Distant Speech Recognition
2021-04-27Virtual Lake Nona Impact Forum “Health Innovation in the New Reality”
2021-04-26FastNeRF: High-Fidelity Neural Rendering at 200FPS [Condensed]
2021-04-21Research for Industries (RFI) Lecture Series: Warren Powell
2021-04-21Research for Industries (RFI) Lecture Series: Andreas Haeberlen
2021-04-13Discovering hidden connections in art with deep, interpretable visual analogies
2021-04-13ZeRO & Fastest BERT: Increasing the scale and speed of deep learning training in DeepSpeed
2021-04-13Interactive sound simulation: Rendering immersive soundscapes in games and virtual reality
2021-04-13A prototype implementation of 4G packet gateway Microsoft Catapult FPGA platform
2021-04-12Self-Tuning Networks: Amortizing the Hypergradient Computation for Hyperparameter Optimization
2021-04-06Ultra-dense data storage and extreme parallelism with electronic-molecular systems