ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)最新文献_第10页

Phase Unwrapping in Correlated Noise for FMCW Lidar Depth Estimation FMCW激光雷达深度估计中相关噪声的相位解包裹

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Pub Date : 2023-06-04 DOI: 10.1109/ICASSP49357.2023.10095456

A. Ulvog, Joshua Rapp, T. Koike-Akino, Hassan Mansour, P. Boufounos, K. Parsons

引用次数: 0

Diffusion-Based Sound Source Localization Using Networks of Planar Microphone Arrays 平面传声器阵列网络中基于扩散的声源定位

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Pub Date : 2023-06-04 DOI: 10.1109/ICASSP49357.2023.10095405

Davide Albertini, Gioele Greco, A. Bernardini, A. Sarti

引用次数: 0

Contrastive Learning-Based Audio to Lyrics Alignment for Multiple Languages 基于对比学习的多语言音频与歌词对齐

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Pub Date : 2023-06-04 DOI: 10.1109/ICASSP49357.2023.10096725

Simon Durand, D. Stoller, Sebastian Ewert

引用次数: 4

High-Speed Drone Detection Based On Yolo-V8 基于Yolo-V8的高速无人机检测

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Pub Date : 2023-06-04 DOI: 10.1109/ICASSP49357.2023.10095516

Jun-Hwa Kim, Namho Kim, C. Won

引用次数: 7

Self-Adaptive Reasoning on Sub-Questions for Multi-Hop Question Answering 多跳问答子问题的自适应推理

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Pub Date : 2023-06-04 DOI: 10.1109/ICASSP49357.2023.10097206

Zekai Li, Wei Peng

{"title":"Self-Adaptive Reasoning on Sub-Questions for Multi-Hop Question Answering","authors":"Zekai Li, Wei Peng","doi":"10.1109/ICASSP49357.2023.10097206","DOIUrl":"https://doi.org/10.1109/ICASSP49357.2023.10097206","url":null,"abstract":"In this paper, we present the Self-Adapting Reasoning Model (SAR) for solving multi-hop question answering (MHQA) tasks, where the QA system is supposed to find the correct answer within the given multiple documents and a multi-hop question. One feasible track on MHQA is question decomposition, based on the idea that a multi-hop question is usually made from several single-hop questions, which are much easier to answer. However, ignoring the inner connection between sub-questions, existing works usually train additional single-hop question-answering models and answer sub-questions separately. To tackle this problem, we design an end-to-end self-adaptive multi-hop reasoning model. Specifically, given a multi-hop question, a question decomposer first decomposes it into two simple questions and identifies the question type. Then, based on the question type, different reasoning strategies are applied for reasoning. This enables our model to be self-adapting and more explainable regarding different types of questions. Experiments are carried out to demonstrate the effectiveness of our model, and SAR achieves remarkable results on the HotpotQA dataset.","PeriodicalId":113072,"journal":{"name":"ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"298 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121454225","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Instance-Aware Hierarchical Structured Policy for Prompt Learning in Vision-Language Models 视觉语言模型中用于快速学习的实例感知分层结构策略

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Pub Date : 2023-06-04 DOI: 10.1109/ICASSP49357.2023.10095231

Xun Wu, Guolong Wang, Zhaoyuan Liu, Xuan Dang, Zheng Qin

{"title":"Instance-Aware Hierarchical Structured Policy for Prompt Learning in Vision-Language Models","authors":"Xun Wu, Guolong Wang, Zhaoyuan Liu, Xuan Dang, Zheng Qin","doi":"10.1109/ICASSP49357.2023.10095231","DOIUrl":"https://doi.org/10.1109/ICASSP49357.2023.10095231","url":null,"abstract":"In recent years, learnable prompts have emerged as a major prompt learning paradigm, enhancing the performance of large-scale vision-language pre-trained models in few-shot image classification. However, enhancing methods are often time-consuming and inflexible because 1) class-specific prompts are inefficient in certain situations; 2) instance-specific prompts are put in a fixed position. To address these issues, inspired by the coarse-to-fine decision-making paradigm of human, we propose an Instance-Aware Hierarchical-Structured Policy (IAHSP) that integrates instance-specific prompt selection and appropriate position selection using a reinforcement learning fashion. Specifically, IAHSP consists of two sub-policies: 1) the root policy selects the most suitable prompt from the prompts pool, and 2) the leaf policy identifies the optimal position for inserting the selected prompt. We train these two policies iteratively with rewards constraining the prompts while maintaining their diversity. Extensive experiments on 11 public benchmarks demonstrate that our IAHSP significantly boosts the few-shot image classification performance of vision-language pre-trained models, while also exhibiting superior generalization performance.","PeriodicalId":113072,"journal":{"name":"ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121718726","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An Adaptive DFE Using Light-Pattern-Protection Algorithm in 12 NM CMOS Technology 基于12纳米CMOS光模式保护算法的自适应DFE

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Pub Date : 2023-06-04 DOI: 10.1109/ICASSP49357.2023.10097007

Shi Xing, Changlong Lin, Yuchen Li, Huandong Wang

引用次数: 0

Augmentation Robust Self-Supervised Learning for Human Activity Recognition 增强鲁棒自监督学习用于人类活动识别

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Pub Date : 2023-06-04 DOI: 10.1109/ICASSP49357.2023.10096151

Cong Xu, Yuhang Li, Dae Lee, Dae Hoon Park, Hongda Mao, Huyen Do, Jonathan Chung, Dinesh Nair

{"title":"Augmentation Robust Self-Supervised Learning for Human Activity Recognition","authors":"Cong Xu, Yuhang Li, Dae Lee, Dae Hoon Park, Hongda Mao, Huyen Do, Jonathan Chung, Dinesh Nair","doi":"10.1109/ICASSP49357.2023.10096151","DOIUrl":"https://doi.org/10.1109/ICASSP49357.2023.10096151","url":null,"abstract":"Human Activity Recognition (HAR) is widely applied on wearable devices in our daily lives. However, acquiring high-quality wearable sensor data set with ground-truths is challenging due to the high cost in collecting data and necessity of domain experts. In order to achieve generalization from limited data, we study augmentation-based Self-Supervised Learning (SSL) for data from wearable devices. However, there is an issue in one of the most popular SSL approaches, contrastive learning: it is sensitive to the choice of data augmentations. To resolve this, we first propose to combine contrastive learning with generative learning, which is robust to augmentations. Second, we propose an automatic augmentation policy search method to discover the most promising augmentation policy. We empirically verify our approaches on three public HAR datasets. Experimental results show that our proposed SSL approach is robust to augmentations, and delivers higher accuracy than contrastive learning. Additionally, with the searched augmentation policy we are able to further improve the accuracy of HAR task.","PeriodicalId":113072,"journal":{"name":"ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121753144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Brain Network Features Differentiate Intentions from Different Emotional Expressions of the Same Text 大脑网络特征区分同一文本不同情感表达的意图

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Pub Date : 2023-06-04 DOI: 10.1109/ICASSP49357.2023.10095376

Zhongjie Li, Bin Zhao, Gaoyan Zhang, J. Dang

{"title":"Brain Network Features Differentiate Intentions from Different Emotional Expressions of the Same Text","authors":"Zhongjie Li, Bin Zhao, Gaoyan Zhang, J. Dang","doi":"10.1109/ICASSP49357.2023.10095376","DOIUrl":"https://doi.org/10.1109/ICASSP49357.2023.10095376","url":null,"abstract":"Intent differentiation in speech communication relies not only on linguistic information but also on paralinguistic information. The same textual content, when pronounced with different prosodies and emotions, may express totally different intentions. The true intentions in this condition can be easily grasped by our brain. Therefore, combining text, speech, and electroencephalography (EEG) for intent discrimination on the same text may be an effective approach. Before fusing speech and text modalities, the current study focused on exploring effective EEG-based features for Chinese intent recognition as no previous research has utilized EEG signals for this purpose. To tackle this issue, we first created a Chinese multimodal spoken language intention understanding (CMSLIU) dataset, in which the same texts were pronounced with varying prosodies to express different intents. To identify effective brain features that were most relevant to intent recognition improvement, we compared the event-related spectral perturbation and effective brain connectivity patterns on two intent conditions (praise vs. irony). It was found that the praise expression tended to elicit stronger high-frequency brain activities while the irony expression involved a more suppressive network connection in the right hemisphere. These features were trained on the CMSLIU dataset and achieved an intention classification accuracy of 78.66%, which indicated a great potential of the EEG features in intent discrimination on the same text.","PeriodicalId":113072,"journal":{"name":"ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-06-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131386922","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Fast Robust Principle Component Analysis Using Gauss-Newton Iterations 基于高斯-牛顿迭代的快速鲁棒主成分分析

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) Pub Date : 2023-06-04 DOI: 10.1109/ICASSP49357.2023.10096269

William Chettleburgh, Zhishen Huang, Ming-Hsuan Yang

引用次数: 0