c. Red Hen Lab - Google Summer of Code 2018 Report
Summer of 2018 was a spectacular year for Red Hen, with twelve projects. A highlight was that for the first time, we achieved an end-to-end pipeline for automatic speech detection of Mandarin Chinese, in production from data capture in China to completed captions in a news dataset of several hundred hours. We also added Russian OCR support for ticker-tape captions, created a smashing Rapid Video Annotator, and made major progress on a series of computer vision projects for emotion detection.
Google Summer of Code 2018 Projects:
Ahmed Ismail -- Automatic Speech Recognition for Arabic -- mentor Ahmed M. H. Abdel-Fattah -- github -- blog
Ajinkya Takawale -- Audio-Visual Speech Recognition -- mentor Karan Singla -- github -- blog
Awani Mishra -- Multimodal Show Segmentation -- mentor Anna Bonazzi -- github
Burhan Ul Tayyab -- Optical Character Recognition in Russian television news shows -- mentor Carlos Fernandez -- github -- blog -- old blog
Devendra Yadav -- Emotion detection and characterization in video using CNN-RNN -- mentor Jakob Suchan -- github -- June slides -- blog
Gyanesh Malhotra -- Multimodal Egocentric Perception, with video, audio, and eye-tracking data -- mentor Tim Groeling
Shuwei Xu -- Automatic Speech Recognition for for Mandarin Chinese -- mentor Kai Chen -- github -- blog
Sumit Vohra -- Temporal Segmentation of Egocentric Videos using audio + videos -- mentor Mehul Bhatt -- github
Vaibhav Gupta -- Red Hen's Rapid Annotator -- mentor Peter Uhrig -- github
Vikrant Goyal -- Multi-language Translation -- mentor Karan Singla -- github -- blog
Vinay Chandragiri -- Emotion Detection -- mentor Francis Steen -- github
Zhaoqing Xu -- Automatic Speech Recognition for for Mandarin Chinese -- mentor Mark Turner -- github -- blog