paper-with-me

홈 › Papers

Factorized Blank Thresholding for Improved Runtime Efficiency of Neural Transducers

2022-11-02 · Duc Le, Frank Seide, Yuhao Wang, Yang Li, Kjell Schubert, Ozlem Kalinli, Michael L. Seltzer

We show how factoring the RNN-T's output distribution can significantly reduce the computation cost and power consumption for on-device ASR inference with no loss in accuracy. With the rise in popularity of neural-transducer type models like the RNN-T for on-device ASR, optimizing RNN-T's runtime efficiency is of great interest. While previous work has primarily focused on the optimization of RNN-T's acoustic encoder and predictor, this paper focuses the attention on the joiner. We show that despite being only a small part of RNN-T, the joiner has a large impact on the overall model's runtime efficiency. We propose to utilize HAT-style joiner factorization for the purpose of skipping the more expensive non-blank computation when the blank probability exceeds a certain threshold. Since the blank probability can be computed very efficiently and the RNN-T output is dominated by blanks, our proposed method leads to a 26-30% decoding speed-up and 43-53% reduction in on-device power consumption, all the while incurring no accuracy degradation and being relatively simple to implement.

📄 PDF Abstract BibTeX arXiv:2211.00896

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Boosting Hybrid Autoregressive Transducer-based ASR with Internal Acoustic Model Training and Dual Blank Thresholding

2024-09-30 · Takafumi Moriya, Takanori Ashihara, Masato Mimura, Hiroshi Sato 외

A hybrid autoregressive transducer (HAT) is a variant of neural transducer that models blank and non-blank posterior distributions separately. In this paper, we propose a novel internal acoustic model (IAM) training stra…

speech-recognitionSpeech Recognition

Effective internal language model training and fusion for factorized transducer model

2024-04-02 · Jinxi Guo, Niko Moritz, Yingyi Ma, Frank Seide 외

The internal language model (ILM) of the neural transducer has been widely studied. In most prior work, it is mainly used for estimating the ILM score and is subsequently subtracted during inference to facilitate improve…

Language ModelingLanguage Modellingmodel

Analyzing the Importance of Blank for CTC-Based Knowledge Distillation

2025-06-02 · Benedikt Hilmes, Nick Rossenbach, Ralf Schlüter

With the rise of large pre-trained foundation models for automatic speech recognition new challenges appear. While the performance of these models is good, runtime and cost of inference increases. One approach to make us…

Automatic Speech RecognitionKnowledge Distillationspeech-recognitionSpeech Recognition

Continuation is a Sub-Task of Fill in the Blank: Why Not Train for Both?

2021-06-16 · ACL ARR Jun 2021 6 · Anonymous

The task of inserting text into a specified position in a passage, known as fill in the blank, is useful for a variety of applications where writers interact with a natural language generation (NLG) system to craft text.…

PositionText Generation

BlanketGen2-Fit3D: Synthetic Blanket Augmentation Towards Improving Real-World In-Bed Blanket Occluded Human Pose Estimation

2025-01-21 · Tamás Karácsony, João Carmona, João Paulo Silva Cunha

Human Pose Estimation (HPE) from monocular RGB images is crucial for clinical in-bed skeleton-based action recognition, however, it poses unique challenges for HPE models due to the frequent presence of blankets occludin…

Action RecognitionData AugmentationPose EstimationSkeleton Based Action Recognition