B
Balakrishnan Varadarajan
Researcher at Google
Publications - 36
Citations - 2221
Balakrishnan Varadarajan is an academic researcher from Google. The author has contributed to research in topics: Hidden Markov model & Semi-supervised learning. The author has an hindex of 15, co-authored 35 publications receiving 1798 citations. Previous affiliations of Balakrishnan Varadarajan include Johns Hopkins University.
Papers
More filters
Posted Content
YouTube-8M: A Large-Scale Video Classification Benchmark
Sami Abu-El-Haija,Nisarg Dilipkumar Kothari,Joonseok Lee,Apostol Natsev,George Toderici,Balakrishnan Varadarajan,Sudheendra Vijayanarasimhan +6 more
TL;DR: YouTube-8M is introduced, the largest multi-label video classification dataset, composed of ~8 million videos (500K hours of video), annotated with a vocabulary of 4800 visual entities, and various (modest) classification models are trained on the dataset.
Posted Content
TNT: Target-driveN Trajectory Prediction
Hang Zhao,Jiyang Gao,Tian Lan,Chen Sun,Benjamin Sapp,Balakrishnan Varadarajan,Yue Shen,Yi Shen,Yuning Chai,Cordelia Schmid,Congcong Li,Dragomir Anguelov +11 more
TL;DR: The key insight is that for prediction within a moderate time horizon, the future modes can be effectively captured by a set of target states, which leads to the target-driven trajectory prediction (TNT) framework.
Journal ArticleDOI
Active learning and semi-supervised learning for speech recognition: A unified framework using the global entropy reduction maximization criterion
TL;DR: It is shown that both the traditional confidence-based active learning and semi-supervised learning approaches can be improved by maximizing the lattice entropy reduction over the whole dataset.
Proceedings ArticleDOI
Unsupervised Learning of Acoustic Sub-word Units
TL;DR: An algorithm for joint unsupervised learning of the topology and parameters of a hidden Markov model (HMM); states and short state-sequences through this HMM correspond to the learnt sub-word units.
Book ChapterDOI
Data-Derived Models for Segmentation with Application to Surgical Assessment and Training
TL;DR: This paper addresses automatic skill assessment in robotic minimally invasive surgery by designed in a data-driven manner, mixing trials from multiple surgeons with varying skill levels, resulting in HMM states that model skill-specific sub-gestures.