ABOUT_INFO = "Polish ASR leaderboard by [AMU-CAI team](https://huggingface.co/amu-cai) aims to provide comprehensive overview of performance of ASR/STT systems for Polish.
\ The leaderboard currently supports [BIGOS V2](https://huggingface.co/datasets/amu-cai/pl-asr-bigos-v2) and [PELCRA for BIGOS](https://huggingface.co/datasets/pelcra/pl-asr-pelcra-for-bigos) datasets.
\ If you want to add your system or dataset to the leaderboard, please contact Michał Junczyk (michal.junczyk@amu.edu.pl) or open a pull request on [GitHub](https://github.com/goodmike31/pl-asr-bigos-tools)
\ To learn more please read blog post [here](https://huggingface.co/blog/michaljunczyk/introducing-polish-asr-leaderboard).
\ If you use this work, please cite it as follows:
\ ```@misc{amu_cai_pl_asr_leaderboard, \ author = {Michał Junczyk}, \ title = {{AMU Polish ASR Leaderboard}}, \ year = {2024}, \ howpublished = {url{https://huggingface.co/spaces/amu-cai/pl-asr-leaderboard}}, \ publisher = {Hugging Face} \ }```" BIGOS_INFO = "BIGOS (Benchmark Intended Grouping of Open Speech) is the collection of freely available speech datasets curated by the [AMU-CAI team](https://huggingface.co/amu-cai). \ Learn more [here](https://huggingface.co/datasets/amu-cai/pl-asr-bigos-v2)" PELCRA_INFO = "PELCRA for BIGOS is the subset of speech corpora created by the [PELCRA group](http://pelcra.pl/new/), curated for the BIGOS benchmark by the [AMU-CAI team](https://huggingface.co/amu-cai). \ Learn more [here](https://huggingface.co/datasets/pelcra/pl-asr-pelcra-for-bigos)" POLEVAL_INFO = "PolEval is test used for Polish ASR challenge. It consists of recordings from BIGOS and PELCRA datasets. For details see: [PolEval 2024 - Task 3 - ASR](https://poleval.pl/tasks/task3)" ANALYSIS_INFO = "Here we examine ASR accuracy depending on the system type, model size, audio duration, speaking rate and speaker charactertics (age and gender)" INSPECTION_INFO = "Here you can inspect the performance of specific ASR systems on the specific audio samples" COMPARISON_INFO = "Here you can compare the performance of different ASR systems on the specific datasets using metrics and visualizations of your choice." asr_systems_colors_mapping = { 'azure': '#1f77b4', # Blue 'google': '#2ca02c', # Green 'wav2vec2': '#d62728', # Red 'nemo': '#9467bd', # Purple 'assemblyai': '#8c564b', # Brown 'mms': '#e377c2', # Pink 'google_v2': '#7f7f7f', # Gray 'whisper_cloud': '#bcbd22', # Olive 'whisper_local': '#ff7f0e', # Orange # Add or override other systems and their colors }