paper-with-me

홈 › Papers

Novel Parasitic Dual-Scale Modeling for Efficient and Accurate Multilingual Speech Translation

2025-08-15 · Chenyang Le, Yinfeng Xia, Huiyan Li, Manhong Wang, Yutao Sun, Xingyang Ma, Yanmin Qian arxiv

Recent advancements in speech-to-text translation have led to the development of multilingual models capable of handling multiple language pairs simultaneously. However, these unified models often suffer from large parameter sizes, making it challenging to balance inference efficiency and performance, particularly in local deployment scenarios. We propose an innovative Parasitic Dual-Scale Approach, which combines an enhanced speculative sampling method with model compression and knowledge distillation techniques. Building on the Whisper Medium model, we enhance it for multilingual speech translation into whisperM2M, and integrate our novel KVSPN module, achieving state-of-the-art (SOTA) performance across six popular languages with improved inference efficiency. KVSPN enables a 40\% speedup with no BLEU score degradation. Combined with distillation methods, it represents a 2.6$\times$ speedup over the original Whisper Medium with superior performance.

📄 PDF Abstract BibTeX arXiv:2508.11189

Code (0)

등록된 구현이 없습니다.

Tasks

Speech-to-Text TranslationKnowledge DistillationModel Compression

Similar Papers 제목 키워드 기반

ParasGB: A Graph Benchmark Suite for Parasitic Estimation on AMS Circuits

2026-07-25 · Jiajun Zou, Jiawei Liu, Ao Liu, Junnong Tian 외 arxiv

As chip manufacturing processes advance to deep submicron nodes, parasitic interconnect effects increasingly dominate the performance of analog and mixed-signal (AMS) circuits and often lead to costly layout iterations. …

Parameter PredictionGraph Learning

Deep-Learning-Based Pre-Layout Parasitic Capacitance Prediction on SRAM Designs

2025-07-09 · Shan Shen, Dingcheng Yang, Yuyang Xie, Chunyan Pei 외 arxiv

To achieve higher system energy efficiency, SRAM in SoCs is often customized. The parasitic effects cause notable discrepancies between pre-layout and post-layout circuit simulations, leading to difficulty in converging …

Graph Neural Network

Mitigate Parasitic Resistance in Resistive Crossbar-based Convolutional Neural Networks

2019-12-17 · Fan Zhang, Miao Hu

Traditional computing hardware often encounters on-chip memory bottleneck on large scale Convolution Neural Networks (CNN) applications. With its unique in-memory computing feature, resistive crossbar-based computing att…

Data-Driven Modeling with Experimental Augmentation for the Modulation Strategy of the Dual-Active-Bridge Converter

2023-07-30 · Xinze Li, Josep Pou, Jiaxin Dong, Fanfan Lin 외

For the performance modeling of power converters, the mainstream approaches are essentially knowledge-based, suffering from heavy manpower burden and low modeling accuracy. Recent emerging data-driven techniques greatly …

End-to-end Malaria Diagnosis and 3D Cell Rendering with Deep Learning

2021-07-08 · Vignav Ramesh

Malaria is a parasitic infection that poses a significant burden on global health. It kills one child every 30 seconds and over one million people annually. If diagnosed in a timely manner, however, most people can be ef…