5 citations · 9 across the 6 of their papers we have counts for
6 papers
Improved Training for End-to-End Streaming Automatic Speech Recognition Model with Punctuation
Hanbyul Kim, Seunghyun Seo, Lukas Lee +1
Punctuated text prediction is crucial for automatic speech recognition as it enhances readability and impacts downstream natural language processing tasks. In streaming scenarios,…
Efficient Storage of Fine-Tuned Models via Low-Rank Approximation of Weight Residuals
Simo Ryu, Seunghyun Seo, Jaejun Yoo
In this paper, we present an efficient method for storing fine-tuned models by leveraging the low-rank properties of weight residuals. Our key observation is that weight residuals…
Neural Attention Memory
Hyoungwook Nam, Seung Byum Seo
We propose a novel perspective of the attention mechanism by reinventing it as a memory architecture for neural networks, namely Neural Attention Memory (NAM). NAM is a memory stru…
Blank Collapse: Compressing CTC emission for the faster decoding
Minkyu Jung, Ohhyeok Kwon, Seunghyun Seo +1
Connectionist Temporal Classification (CTC) model is a very efficient method for modeling sequences, especially for speech data. In order to use CTC model as an Automatic Speech Re…
Integration of Pre-trained Networks with Continuous Token Interface for End-to-End Spoken Language Understanding
Seunghyun Seo, Donghyun Kwak, Bowon Lee
Most End-to-End (E2E) SLU networks leverage the pre-trained ASR networks but still lack the capability to understand the semantics of utterances, crucial for the SLU task. To solve…
I-BERT: Inductive Generalization of Transformer to Arbitrary Context Lengths
Hyoungwook Nam, Seung Byum Seo, Vikram Sharma Mailthody +2
Self-attention has emerged as a vital component of state-of-the-art sequence-to-sequence models for natural language processing in recent years, brought to the forefront by pre-tra…