--- dataset_info: - config_name: audio/FLEURS/assamese features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 400129016.0 num_examples: 984 download_size: 394387983 dataset_size: 400129016.0 - config_name: audio/FLEURS/bengali features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 395894197.0 num_examples: 920 download_size: 395008857 dataset_size: 395894197.0 - config_name: audio/FLEURS/gujarati features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 336184742.0 num_examples: 1000 download_size: 334871945 dataset_size: 336184742.0 - config_name: audio/FLEURS/hindi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 154790380.0 num_examples: 418 download_size: 147578725 dataset_size: 154790380.0 - config_name: audio/FLEURS/kannada features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 366376573.0 num_examples: 838 download_size: 358502243 dataset_size: 366376573.0 - config_name: audio/FLEURS/malayalam features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 450770445.0 num_examples: 958 download_size: 441005995 dataset_size: 450770445.0 - config_name: audio/FLEURS/marathi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 443738065.13 num_examples: 1015 download_size: 436518279 dataset_size: 443738065.13 - config_name: audio/FLEURS/nepali features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 263454706.0 num_examples: 726 download_size: 258043059 dataset_size: 263454706.0 - config_name: audio/FLEURS/odia features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 341961593.0 num_examples: 883 download_size: 320745382 dataset_size: 341961593.0 - config_name: audio/FLEURS/punjabi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 211791247.0 num_examples: 574 download_size: 205760432 dataset_size: 211791247.0 - config_name: audio/FLEURS/sindhi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 379527787.0 num_examples: 980 download_size: 377615930 dataset_size: 379527787.0 - config_name: audio/FLEURS/tamil features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 245793399.0 num_examples: 591 download_size: 240410513 dataset_size: 245793399.0 - config_name: audio/FLEURS/telugu features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 167107575.0 num_examples: 472 download_size: 163463460 dataset_size: 167107575.0 - config_name: audio/FLEURS/urdu features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 93907962.0 num_examples: 299 download_size: 93516805 dataset_size: 93907962.0 - config_name: audio/commonvoice/assamese features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 56283901.0 num_examples: 308 download_size: 48754638 dataset_size: 56283901.0 - config_name: audio/commonvoice/bengali features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 1904104090.168 num_examples: 9168 download_size: 1684078386 dataset_size: 1904104090.168 - config_name: audio/commonvoice/hindi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 513842146.6 num_examples: 2962 download_size: 435100874 dataset_size: 513842146.6 - config_name: audio/commonvoice/malayalam features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 17282290.0 num_examples: 146 download_size: 16021243 dataset_size: 17282290.0 - config_name: audio/commonvoice/marathi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 375551010.368 num_examples: 1827 download_size: 333389999 dataset_size: 375551010.368 - config_name: audio/commonvoice/nepali features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 9193085.0 num_examples: 66 download_size: 8538576 dataset_size: 9193085.0 - config_name: audio/commonvoice/odia features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 39990835.0 num_examples: 226 download_size: 33675531 dataset_size: 39990835.0 - config_name: audio/commonvoice/punjabi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 60745731.0 num_examples: 414 download_size: 55805591 dataset_size: 60745731.0 - config_name: audio/commonvoice/tamil features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 2259861523.16 num_examples: 11955 download_size: 1991341063 dataset_size: 2259861523.16 - config_name: audio/commonvoice/urdu features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 466426558.096 num_examples: 3301 download_size: 407722998 dataset_size: 466426558.096 - config_name: audio/gramvaani/hindi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 322153519.264 num_examples: 1032 download_size: 320173679 dataset_size: 322153519.264 - config_name: audio/indictts/bengali features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 17827661.0 num_examples: 100 download_size: 17262297 dataset_size: 17827661.0 - config_name: audio/indictts/gujarati features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 40820114.0 num_examples: 100 download_size: 39721020 dataset_size: 40820114.0 - config_name: audio/indictts/hindi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 24242247.0 num_examples: 100 download_size: 22378518 dataset_size: 24242247.0 - config_name: audio/indictts/kannada features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 25119550.0 num_examples: 100 download_size: 23441314 dataset_size: 25119550.0 - config_name: audio/indictts/malayalam features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 18528262.0 num_examples: 100 download_size: 17141988 dataset_size: 18528262.0 - config_name: audio/indictts/marathi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 23234538.0 num_examples: 100 download_size: 20049326 dataset_size: 23234538.0 - config_name: audio/indictts/odia features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 18500622.0 num_examples: 100 download_size: 16381711 dataset_size: 18500622.0 - config_name: audio/indictts/tamil features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 35304113.0 num_examples: 100 download_size: 33130701 dataset_size: 35304113.0 - config_name: audio/indictts/telugu features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 48953745.0 num_examples: 100 download_size: 45226295 dataset_size: 48953745.0 - config_name: audio/kathbath/bengali features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 366265573.86 num_examples: 1783 download_size: 361373961 dataset_size: 366265573.86 - config_name: audio/kathbath/gujarati features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 349708715.498 num_examples: 1766 download_size: 344143519 dataset_size: 349708715.498 - config_name: audio/kathbath/kannada features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 358809666.14 num_examples: 1388 download_size: 344241499 dataset_size: 358809666.14 - config_name: audio/kathbath/marathi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 350433007.579 num_examples: 1631 download_size: 345696700 dataset_size: 350433007.579 - config_name: audio/kathbath/odia features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 372088901.96 num_examples: 1862 download_size: 346120350 dataset_size: 372088901.96 - config_name: audio/kathbath/punjabi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 369255097.688 num_examples: 1914 download_size: 335986442 dataset_size: 369255097.688 - config_name: audio/kathbath/sanskrit features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 347328171.57 num_examples: 1090 download_size: 340063647 dataset_size: 347328171.57 - config_name: audio/kathbath/tamil features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 349049035.286 num_examples: 1642 download_size: 359656982 dataset_size: 349049035.286 - config_name: audio/kathbath/telugu features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 364214162.172 num_examples: 1492 download_size: 349947735 dataset_size: 364214162.172 - config_name: audio/kathbath/urdu features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 368062421.574 num_examples: 1959 download_size: 347304086 dataset_size: 368062421.574 - config_name: audio/mucs/gujarati features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 666434960.784 num_examples: 3419 download_size: 567438725 dataset_size: 666434960.784 - config_name: audio/mucs/hindi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 635930603.68 num_examples: 3897 download_size: 607639615 dataset_size: 635930603.68 - config_name: audio/mucs/marathi features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 77024257.0 num_examples: 636 download_size: 75975181 dataset_size: 77024257.0 - config_name: audio/mucs/odia features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 564160357.36 num_examples: 4420 download_size: 519258461 dataset_size: 564160357.36 - config_name: audio/mucs/tamil features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 523709556.064 num_examples: 2609 download_size: 500876199 dataset_size: 523709556.064 - config_name: audio/mucs/telugu features: - name: audio dtype: audio - name: language dtype: string - name: transcript dtype: string splits: - name: train num_bytes: 495266332.124 num_examples: 2549 download_size: 494383639 dataset_size: 495266332.124 configs: - config_name: FLEURS_assamese data_files: - split: train path: audio/FLEURS/assamese/train-* - config_name: FLEURS_bengali data_files: - split: train path: audio/FLEURS/bengali/train-* - config_name: FLEURS_gujarati data_files: - split: train path: audio/FLEURS/gujarati/train-* - config_name: FLEURS_hindi data_files: - split: train path: audio/FLEURS/hindi/train-* - config_name: FLEURS_kannada data_files: - split: train path: audio/FLEURS/kannada/train-* - config_name: FLEURS_malayalam data_files: - split: train path: audio/FLEURS/malayalam/train-* - config_name: FLEURS_marathi data_files: - split: train path: audio/FLEURS/marathi/train-* - config_name: FLEURS_nepali data_files: - split: train path: audio/FLEURS/nepali/train-* - config_name: FLEURS_odia data_files: - split: train path: audio/FLEURS/odia/train-* - config_name: FLEURS_punjabi data_files: - split: train path: audio/FLEURS/punjabi/train-* - config_name: FLEURS_sindhi data_files: - split: train path: audio/FLEURS/sindhi/train-* - config_name: FLEURS_tamil data_files: - split: train path: audio/FLEURS/tamil/train-* - config_name: FLEURS_telugu data_files: - split: train path: audio/FLEURS/telugu/train-* - config_name: FLEURS_urdu data_files: - split: train path: audio/FLEURS/urdu/train-* - config_name: commonvoice_assamese data_files: - split: train path: audio/commonvoice/assamese/train-* - config_name: commonvoice_bengali data_files: - split: train path: audio/commonvoice/bengali/train-* - config_name: commonvoice_hindi data_files: - split: train path: audio/commonvoice/hindi/train-* - config_name: commonvoice_malayalam data_files: - split: train path: audio/commonvoice/malayalam/train-* - config_name: commonvoice_marathi data_files: - split: train path: audio/commonvoice/marathi/train-* - config_name: commonvoice_nepali data_files: - split: train path: audio/commonvoice/nepali/train-* - config_name: commonvoice_odia data_files: - split: train path: audio/commonvoice/odia/train-* - config_name: commonvoice_punjabi data_files: - split: train path: audio/commonvoice/punjabi/train-* - config_name: commonvoice_tamil data_files: - split: train path: audio/commonvoice/tamil/train-* - config_name: commonvoice_urdu data_files: - split: train path: audio/commonvoice/urdu/train-* - config_name: gramvaani_hindi data_files: - split: train path: audio/gramvaani/hindi/train-* - config_name: indictts_bengali data_files: - split: train path: audio/indictts/bengali/train-* - config_name: indictts_gujarati data_files: - split: train path: audio/indictts/gujarati/train-* - config_name: indictts_hindi data_files: - split: train path: audio/indictts/hindi/train-* - config_name: indictts_kannada data_files: - split: train path: audio/indictts/kannada/train-* - config_name: indictts_malayalam data_files: - split: train path: audio/indictts/malayalam/train-* - config_name: indictts_marathi data_files: - split: train path: audio/indictts/marathi/train-* - config_name: indictts_odia data_files: - split: train path: audio/indictts/odia/train-* - config_name: indictts_tamil data_files: - split: train path: audio/indictts/tamil/train-* - config_name: indictts_telugu data_files: - split: train path: audio/indictts/telugu/train-* - config_name: kathbath_bengali data_files: - split: train path: audio/kathbath/bengali/train-* - config_name: kathbath_gujarati data_files: - split: train path: audio/kathbath/gujarati/train-* - config_name: kathbath_kannada data_files: - split: train path: audio/kathbath/kannada/train-* - config_name: kathbath_marathi data_files: - split: train path: audio/kathbath/marathi/train-* - config_name: kathbath_odia data_files: - split: train path: audio/kathbath/odia/train-* - config_name: kathbath_punjabi data_files: - split: train path: audio/kathbath/punjabi/train-* - config_name: kathbath_sanskrit data_files: - split: train path: audio/kathbath/sanskrit/train-* - config_name: kathbath_tamil data_files: - split: train path: audio/kathbath/tamil/train-* - config_name: kathbath_telugu data_files: - split: train path: audio/kathbath/telugu/train-* - config_name: kathbath_urdu data_files: - split: train path: audio/kathbath/urdu/train-* - config_name: mucs_gujarati data_files: - split: train path: audio/mucs/gujarati/train-* - config_name: mucs_hindi data_files: - split: train path: audio/mucs/hindi/train-* - config_name: mucs_marathi data_files: - split: train path: audio/mucs/marathi/train-* - config_name: mucs_odia data_files: - split: train path: audio/mucs/odia/train-* - config_name: mucs_tamil data_files: - split: train path: audio/mucs/tamil/train-* - config_name: mucs_telugu data_files: - split: train path: audio/mucs/telugu/train-* --- # Vaani ASR Benchmark: Comprehensive Evaluation of Indian Language Speech Recognition ## About the Vaani ASR Benchmark The **Vaani ASR Benchmark** is a comprehensive evaluation framework designed to assess the performance of Automatic Speech Recognition (ASR) models across multiple Indian languages. This benchmark addresses the critical need for standardized evaluation of ASR systems in the linguistically diverse Indian subcontinent, where over 700 languages are spoken with 22 official languages recognized by the Constitution. ### Why This Benchmark Matters **Addressing the Indian Language Gap**: While significant progress has been made in ASR for high-resource languages like English and Mandarin, Indian languages have remained underrepresented in speech recognition research. The Vaani benchmark fills this critical gap by providing: - **Standardized Evaluation**: Consistent metrics and methodology across different models and languages - **Diverse Linguistic Coverage**: Support for major Indian languages including Hindi, Tamil, Telugu, Kannada, Bengali, and more - **Real-world Applicability**: Evaluation datasets that reflect actual usage scenarios across India - **Research Acceleration**: A common platform for researchers to compare and improve their ASR models ### What We Evaluate The benchmark evaluates ASR models across multiple dimensions: **🎯 Primary Metrics** - **Word Error Rate (WER)**: Percentage of words incorrectly recognized (lower is better) - **Character Error Rate (CER)**: Percentage of characters incorrectly recognized (lower is better) **📊 Multiple Test Sets** Our evaluation incorporates diverse, high-quality datasets: 1. **FLEURS (Google)**: Multilingual speech corpus with 102 languages, providing ~10 hours per language with parallel sentences for robust cross-linguistic evaluation 2. **Common Voice 12.0 (Mozilla)**: Community-contributed dataset with 26,119+ recorded hours across 104 languages, including rich demographic metadata (age, gender, accent) 3. **IndicVoices (AI4Bharat)**: 12,000 hours of natural Indian speech covering 22 languages with diverse content: - Read speech (8%) - Extempore speech (76%) - Conversational speech (15%) - 22,563 speakers across 208 Indian districts 4. **Gramvaani Hindi Dataset**: Specialized Hindi ASR benchmark focusing on agriculture, healthcare, and general knowledge domains 5. **MUCS 2021**: Multilingual and code-switching dataset with ~600 hours across 7 Indian languages, including Hindi-English and Bengali-English code-switching 6. **IndicTTS Database**: 10,000+ utterances per language across 22 Indian languages with both native and English content 7. **Kathbath (IndicSUPERB)**: 1,684 hours of labeled speech data across 12 Indian languages for comprehensive speech understanding evaluation ### How We Evaluate **🔬 Rigorous Methodology** Our evaluation follows a standardized protocol ensuring fair and accurate assessment: **Text Preprocessing Pipeline:** ```python def clean(text): # Remove annotations and markup text = re.sub(r'{[^}]*}','',text) # Remove {annotations} text = re.sub("[([].*?[)]]", "", text) # Remove [brackets] and (parentheses) text = re.sub('<[^>]+>', '', text) # Remove HTML/XML tags # Normalize punctuation text = text.replace("।", " ").replace("|", " ").replace("-", " ")\ .replace(".", " ").replace(",", " ").replace("I", " ")\ .replace('\n', ' ') # Normalize spacing text = re.sub(' +', ' ', text) return text.strip() ``` **Error Rate Calculation:** - Uses industry-standard `jiwer` library for accurate WER/CER computation - Identical preprocessing applied to both reference and hypothesis texts - Results scaled to percentage (0-100) with 2-decimal precision - Handles edge cases and missing data appropriately ### Language Coverage **🗣️ Multilingual Support** The benchmark currently supports major Indian languages with plans for expansion: **Currently Supported:** - **Indo-Aryan**: Hindi, Bengali, Marathi, Gujarati, Punjabi, Urdu, Assamese, Odia, Nepali - **Dravidian**: Tamil, Telugu, Kannada, Malayalam - **Tibeto-Burman**: Manipuri, Bodo - **Others**: Sanskrit, Santhali **Planned Expansion:** - Additional regional languages and dialects - Tribal and minority languages - Code-switching scenarios (Hindi-English, Tamil-English, etc.) ### Dataset Characteristics **📈 Comprehensive Coverage** Our test datasets provide diverse evaluation scenarios: **Audio Quality Spectrum:** - Studio-quality recordings for controlled evaluation - Real-world recordings capturing natural speech variations - Telephonic and mobile recordings for practical applications **Speaker Diversity:** - **Demographics**: Balanced age, gender, and regional representation - **Accents**: Multiple dialectal variations within languages - **Speaking Styles**: Read speech, spontaneous speech, conversational audio **Content Variety:** - **Domains**: News, agriculture, healthcare, education, general knowledge - **Speech Types**: Formal presentations, casual conversations, prompted responses - **Acoustic Conditions**: Clean studio, noisy environments, multiple speakers ### Performance Analysis **📊 Detailed Metrics** - **AVG WER/CER**: Simple average across all test datasets - **Language-specific Performance**: Individual language breakdowns - **Dataset-specific Analysis**: Performance variations across different test sets - **Statistical Significance**: Confidence intervals and significance testing **🔍 Interactive Exploration** - **Metric Selector**: Switch between WER and CER views - **Language Filtering**: Focus on specific languages or language families - **Dataset Comparison**: Compare model performance across different test sets - **Trend Analysis**: Track model improvements over time ### Research Impact **🎯 Advancing Indian Language ASR** The Vaani benchmark serves multiple stakeholders: **For Researchers:** - Standardized evaluation platform for model comparison - Comprehensive datasets for training and testing - Open-source evaluation code for reproducibility **For Industry:** - Performance benchmarks for commercial ASR systems - Quality assurance metrics for product development - Market readiness assessment for Indian language applications **For Society:** - Enabling voice interfaces in local languages - Supporting digital inclusion across linguistic communities - Preserving and promoting linguistic diversity through technology ### Technical Implementation **🛠️ Robust Infrastructure** - **Scalable Evaluation**: Automated pipeline handling large-scale model evaluation - **Reproducible Results**: Version-controlled datasets and evaluation scripts - **Quality Assurance**: Multiple validation checkpoints and error detection - **Open Source**: Full transparency in methodology and implementation ### Future Roadmap **🚀 Continuous Enhancement** - **Dataset Expansion**: Adding more languages and domains - **Metric Refinement**: Incorporating semantic and contextual evaluation measures - **Real-time Evaluation**: Support for streaming ASR model assessment - **Community Integration**: Enabling community contributions and model submissions --- ## Citation If you use this benchmark in your research, please cite: ```bibtex @misc{vaani_asr_benchmark_2024, title={Vaani ASR Benchmark: Comprehensive Evaluation Framework for Indian Language Speech Recognition}, author={Vaani Team}, year={2024}, url={https://vaani.iisc.ac.in} } ``` For individual datasets used in the benchmark, please also cite the original sources as provided in our dataset documentation.