28 citations · 50 across the 13 of their papers we have counts for
14 papers
From Masking to Merging: Rethinking SpecAugment for Efficient Audio Spectrogram Transformer
Minhee Park, Hyowon Ahn, Chanwoo Kim
This paper proposes SpecAugment-Patch Merging, a simple yet effective method to accelerate Audio Spectrogram Transformer (AST) training. We first apply SpecAugment to mask input sp…
Macro-block dropout for improved regularization in training end-to-end speech recognition models
Chanwoo Kim, Sathish Indurti, Jinhwan Park +1
This paper proposes a new regularization algorithm referred to as macro-block dropout. The overfitting issue has been a difficult problem in training large neural network models. T…
Transformer-based Global 3D Hand Pose Estimation in Two Hands Manipulating Objects Scenarios
Hoseong Cho, Donguk Kim, Chanwoo Kim +2
This report describes our 1st place solution to ECCV 2022 challenge on Human Body, Hands, and Activities (HBHA) from Egocentric and Multi-view Cameras (hand pose estimation). In th…
A comparison of streaming models and data augmentation methods for robust speech recognition
Jiyeon Kim, Mehul Kumar, Dhananjaya Gowda +2
In this paper, we present a comparative study on the robustness of two different online streaming speech recognition models: Monotonic Chunkwise Attention (MoChA) and Recurrent Neu…
Streaming end-to-end speech recognition with jointly trained neural feature enhancement
Chanwoo Kim, Abhinav Garg, Dhananjaya Gowda +2
In this paper, we present a streaming end-to-end speech recognition model based on Monotonic Chunkwise Attention (MoCha) jointly trained with enhancement layers. Even though the Mo…
Faster Re-translation Using Non-Autoregressive Model For Simultaneous Neural Machine Translation
Hyojung Han, Sathish Indurthi, Mohd Abbas Zaidi +5
Recently, simultaneous translation has gathered a lot of attention since it enables compelling applications such as subtitle translation for a live event or real-time video-call tr…
A review of on-device fully neural end-to-end automatic speech recognition algorithms
Chanwoo Kim, Dhananjaya Gowda, Dongsoo Lee +5
In this paper, we review various end-to-end automatic speech recognition algorithms and their optimization techniques for on-device applications. Conventional speech recognition sy…
Small energy masking for improved neural network training for end-to-end speech recognition
Chanwoo Kim, Kwangyoun Kim, Sathish Reddy Indurthi
In this paper, we present a Small Energy Masking (SEM) algorithm, which masks inputs having values below a certain threshold. More specifically, a time-frequency bin is masked if t…
Attention based on-device streaming speech recognition with large speech corpus
Kwangyoun Kim, Kyungmin Lee, Dhananjaya Gowda +10
In this paper, we present a new on-device automatic speech recognition (ASR) system based on monotonic chunk-wise attention (MoChA) models trained with large (> 10K hours) corpus.…
Improved Multi-Stage Training of Online Attention-based Encoder-Decoder Models
Abhinav Garg, Dhananjaya Gowda, Ankur Kumar +3
In this paper, we propose a refined multi-stage multi-task training strategy to improve the performance of online attention-based encoder-decoder (AED) models. A three-stage traini…
power-law nonlinearity with maximally uniform distribution criterion for improved neural network training in automatic speech recognition
Chanwoo Kim, Mehul Kumar, Kwangyoun Kim +1
In this paper, we describe the Maximum Uniformity of Distribution (MUD) algorithm with the power-law nonlinearity. In this approach, we hypothesize that neural network training wil…
end-to-end training of a large vocabulary end-to-end speech recognition system
Chanwoo Kim, Sungsoo Kim, Kwangyoun Kim +10
In this paper, we present an end-to-end training framework for building state-of-the-art end-to-end speech recognition systems. Our training system utilizes a cluster of Central Pr…
Data Efficient Direct Speech-to-Text Translation with Modality Agnostic Meta-Learning
Sathish Indurthi, Houjeung Han, Nikhil Kumar Lakumarapu +4
End-to-end Speech Translation (ST) models have several advantages such as lower latency, smaller model size, and less error compounding over conventional pipelines that combine Aut…
Efficient Implementation of the Room Simulator for Training Deep Neural Network Acoustic Models
Chanwoo Kim, Ehsan Variani, Arun Narayanan +1
In this paper, we describe how to efficiently implement an acoustic room simulator to generate large-scale simulated data for training deep neural networks. Even though Google Room…