Webb除了比赛提供的数据,还用到了SLR17(加性噪声)/SLR28(混响噪声)两个噪声库和一个高信噪比的带文本标注的语音数据SLR33。 用一系列数据增强的方法扩充训练数据到200万 用aishell1训练声学模型提供对齐信息,改进kaldi基于能量的VAD 对齐后就可以知道每一帧是否是Speech,是不是要保留 对比各种预处理方式的效果 模型方面使用了三个 i-vector … Webb20 aug. 2024 · 2.SLR33 Aishell. Aishell is an open-source Chinese Mandarin speech corpus published by Beijing Shell Shell Technology Co.,Ltd. 400 people from different accent areas in China are invited to participate in the recording, which is conducted in a quiet indoor environment using high fidelity microphone and downsampled to 16kHz.
希尔贝壳—专注于人工智能大数据和技术的创新
WebbAISHELL-4是一个通过麦克风阵列实录的八通道中文普通话会议场景语音数据集。. 该数据集共包含211场会议,每场会议4至8人,数据集共120小时左右。. 该数据集旨在促进实际应用场景下多说话人处理的研究。. AISHELL-4数据包括了实际会议场景下各种重要特性,例如 ... WebbAirshells. 5,157 likes · 5 talking about this. Airshells protects your stroller, bike or wheelchair during travel, to make sure you get the best po church law lexington
AISHELL-3 Baseline Samples - GitHub Pages
WebbIf you want use my aishell dataset code, you also should take care about the transcripts file path in data/aishell.py line 26: src_file = "/data/Speech/SLR33/data_aishell/" + "transcript/aishell_transcript_v0.8.txt" When ready. Let's train: python main.py --config ./config/aishell_asr_example_lstm4atthead1.yaml WebbCannot retrieve contributors at this time. 58 lines (56 sloc) 2.35 KB. Raw Blame. data: corpus: name: 'aishell' # Specify dataset. path: '/data/Speech/SLR33/data_aishell/wav/' # Path to raw LibriSpeech dataset. train_split: ['train'] # … Webb7 mars 2024 · 2.SLR33 Aishell Aishell is an open-source Chinese Mandarin speech corpus published by Beijing Shell Shell Technology Co.,Ltd. 400 people from different accent areas in China are invited to participate in the recording, which is conducted in a quiet indoor environment using high fidelity microphone and downsampled to 16kHz. dewalt battery tester