If you use language model, you need to install the KenLM bindings with: conda activate your_environment If you want to prepare a database in HuggingFace format, you can follow the data loader script in: data_loader_atc.py.However, do not worry, we have prepared the database in Datasets format:.We use the UWB-ATCC + ATCOSIM corpus to fine-tune this model.
#300m atc code 24 how to
We described there the partitions of how to use our model. See Table 1 (page 3) in our paper: How Does Pre-trained Wav2Vec 2.0 Perform on Domain Shifted ASR? An Extensive Benchmark on Air Traffic Control Communications. We don't expect that it keeps the same performance on some others datasets, e.g., LibriSpeech or CommonVoice. This model was fine-tuned on air traffic control data. You can use our Google Colab notebook to run and evaluate our model: We analyze WERs on the low-resource scenario and gender bias carried by one ATC dataset. Relative word error rate (WER) reductions between 20% to 40% are obtained in comparison to hybrid-based ASR baselines by only fine-tuning E2E acoustic models with a smaller fraction of labeled data. We benchmark these two models on several open-source and challenging ATC databases with signal-to-noise ratio between 5 and 20 dB. We target this scenario by analyzing the robustness of Wav2Vec 2.0 and XLS-R models on downstream ASR for a completely unseen domain, air traffic control (ATC) communications. Yet, few works investigated the impact on performance when the data properties substantially differ between the pre-training and fine-tuning phases, termed domain shift. Paper: How Does Pre-trained Wav2Vec 2.0 Perform on Domain Shifted ASR? An Extensive Benchmark on Air Traffic Control Communications.Īuthors: Juan Zuluaga-Gomez, Amrutha Prasad, Iuliia Nigmatulina, Saeed Sarfjoo, Petr Motlicek, Matthias Kleinert, Hartmut Helmke, Oliver Ohneiser, Qingran ZhanĪbstract: Recent work on self-supervised pre-training focus on leveraging large-scale unlabeled speech data to build robust end-to-end (E2E)acoustic models (AM) that can be later fine-tuned on downstream tasks e.g., automatic speech recognition (ASR). It achieves the following results on the evaluation set (two tests sets joined together: UWB-ATCC and ATCOSIM): This model is a fine-tuned version of facebook/wav2vec2-xls-r-300m on two corpus: Wav2vec2-xls-r-300m-en-atc-uwb-atcc-and-atcosim
0 kommentar(er)
