CoSAm
收藏资源简介:
CoSAm数据集由印度德拉敦的UPES、英国伦敦德里的Ulster和印度德里的IIIT-Delhi联合创建,专注于自闭症谱系障碍(ASD)儿童的代码切换(英语和印地语)语音分析。该数据集包含61条语音记录,总计159.75分钟,来自30名ASD儿童和31名神经正常儿童。数据集的创建过程包括从ASD儿童和对照组中收集语音记录,并进行预处理和特征分析。CoSAm数据集主要应用于ASD的早期检测和诊断,通过分析语音特征来提高识别的准确性和鲁棒性。
The CoSAm dataset was jointly developed by UPES in Dehradun, India, Ulster in Derry, UK, and IIIT-Delhi in Delhi, India. It focuses on code-switching (English and Hindi) speech analysis for children with Autism Spectrum Disorder (ASD). This dataset comprises 61 speech recordings totaling 159.75 minutes, collected from 30 children with ASD and 31 neurotypical children. The development process of the CoSAm dataset includes collecting speech recordings from children with ASD and the control group, followed by preprocessing and feature analysis. The CoSAm dataset is primarily applied to the early detection and diagnosis of ASD, aiming to improve the accuracy and robustness of recognition via speech feature analysis.

- 1Modality-Order Matters! A Novel Hierarchical Feature Fusion Method for CoSAm: A Code-Switched Autism CorpusUPES, Bidholi road, Dehradun, 248001, India; Ulster, Northland Road, Londonderry, BT1 6DN., United Kingdom; IIIT-Delhi, Okhla Road, 695571, Delhi, India · 2024年



