AIConfPaper
← Search

Somshubra Majumdar

6 accepted papers

2025

SWAN: An Efficient and Scalable Approach for Long-Context Language Modeling

Krishna C Puvvada, Faisal Ladhak, Santiago Akle Serano, Cheng-Ping Hsieh, Shantanu Acharya, Somshubra Majumdar +5
EMNLP 2025

We present SWAN, a causal Transformer architecture in the decoder-only style that generalizes robustly to sequence lengths substantially longer than those seen during training. SWAN interleaves layers without positional encodings (NoPE) and sliding-window attention layers equipped with rotary positi

Cited by 0Source ↗PDF ↗code ↗Scholar ↗
2024

Investigating End-to-End ASR Architectures for Long Form Audio Transcription

Nithin Rao Koluguri, Samuel Kriman, Georgy Zelenfroind, Somshubra Majumdar, Dima Rekesh, Vahid Noroozi +2
ICASSP 2024accepted

This paper presents an overview and evaluation of some of the end-to-end ASR models on long-form audio. We study three categories of Automatic Speech Recognition(ASR) models based on their core architecture: (1) convolutional, (2) convolutional with squeeze-and-excitation, and (3) convolutional mode…

Cited by 0Source ↗Scholar ↗
2024

Stateful Conformer with Cache-Based Inference for Streaming Automatic Speech Recognition

Vahid Noroozi, Somshubra Majumdar, Ankur Kumar, Jagadeesh Balam, Boris Ginsburg
ICASSP 2024accepted

In this paper, we propose an efficient and accurate streaming speech recognition model based on the FastConformer architecture. We adapted the FastConformer architecture for streaming applications through: (1) constraining both the look-ahead and past contexts in the encoder, and (2) introducing an…

Cited by 0Source ↗Scholar ↗
2023

Efficient Sequence Transduction by Jointly Predicting Tokens and Durations

Hainan Xu, Fei Jia, Somshubra Majumdar, He Huang, Shinji Watanabe, Boris Ginsburg
ICML 2023poster

This paper introduces a novel Token-and-Duration Transducer (TDT) architecture for sequence-to-sequence tasks. TDT extends conventional RNN-Transducer architectures by jointly predicting both a token and its duration, i.e. the number of input frames covered by the emitted token. This is achieved by…

Cited by 15Source ↗PDF ↗code ↗Scholar ↗
2023

Multi-Blank Transducers for Speech Recognition

Hainan Xu, Fei Jia, Somshubra Majumdar, Shinji Watanabe, Boris Ginsburg
ICASSP 2023accepted

This paper proposes a modification to RNN-Transducer (RNN-T) models for automatic speech recognition (ASR). In standard RNN-T, the emission of a blank symbol consumes exactly one input frame; in our proposed method, we introduce additional blank symbols, which consume two or more input frames when e…

Cited by 0Source ↗Scholar ↗
2021

MarbleNet: Deep 1D Time-Channel Separable Convolutional Neural Network for Voice Activity Detection

Fei Jia, Somshubra Majumdar, Boris Ginsburg
ICASSP 2021accepted

We present MarbleNet, an end-to-end neural network for Voice Activity Detection (VAD). MarbleNet is a deep residual network composed from blocks of 1D time-channel separable convolution, batch-normalization, ReLU and dropout layers. When compared to a state-of-the-art VAD model, MarbleNet is able to…

Cited by 0Source ↗Scholar ↗