Kashmiri speech recognition resources
Recorded Kashmiri speech with transcripts, and models that turn spoken Kashmiri into text.
- 25
- Resources2017–2026
- 13
- Datasets
- 3
- Models
- 5
- Papers
Filter this task in the full hub · Guide: Kashmiri language data · Kashmiri Speech Technology: Datasets for TTS and Speech Recognition
Datasets 13
100 Shruk Speech-to-Text Dataset
Recordings of traditional Kashmiri shruks paired with Kashmiri-script transcriptions and thematic tags (99 samples), for training and evaluating Kashmiri ASR.
Chaashini
Continuously growing corpus of clean, single-speaker, studio-grade speech in 33 Indian languages (about 2,955 hours), each clip passed through a multi-stage quality gate. Includes Kashmiri.
Indic DiarBench
Joint diarization and ASR benchmark: about 108 hours of natural multi-speaker conversation across the 22 scheduled languages, with human-corrected, speaker-attributed transcripts. Includes Kashmiri.
Kashmiri Audio Corpus (Segmented)
1,955 segmented speech samples (16kHz, 16-bit, mono WAV) derived from OpenSLR-122. Ready-to-use format for ASR training pipelines.
Omnilingual ASR Corpus
Spontaneous speech recordings with transcriptions in 348 under-served languages, collected for Meta's Omnilingual ASR project. Includes Kashmiri (kas_Arab) train, dev and test splits.
UrduSpeech — Urdu & Kashmiri Speech Corpus
Multi-style, professionally recorded speech corpus covering Urdu and Kashmiri, designed for both text-to-speech and automatic speech recognition.
IndicVoices — Kashmiri
AI4Bharat's natural and spontaneous speech corpus for 22 Indian languages (23.7K hours in total), with a Kashmiri split for speech recognition.
MMS ulab v2
8,900 hours of unlabelled speech in 4,023 languages for language identification, spoken language modelling and representation learning; a reproduced and extended MMS ulab.
Indian Kashmiri Speech Recognition Corpus (Mobile)
Commercial ASR corpus: 115.34 hours of Kashmiri speech from 218 speakers (102 male, 116 female) across 77,255 utterances recorded in quiet environments. The largest Kashmiri speech collection known, though it is licensed rather than open.
Kashmiri Raw Speech Corpus (LDC-IL)
Natural raw speech corpus for Kashmiri covering multiple dialects, age groups and both genders. Speech was collected across the Valley from Pulwama, Srinagar and Anantnag, comprising words, sentences and running text.
Kashmiri Data Corpus (OpenSLR-122)
Transcribed audio recordings from native Kashmiri speakers for Automatic Speech Recognition (ASR) development. The foundational Kashmiri speech dataset. GPL-3.0 licensed.
Kashmiri Spoken Words Dataset
Processed audio data of 12 frequently spoken Kashmiri words. Designed for spoken word recognition and voice command research.
kscp — Kashmiri Speech Corpus Processing
Tools for processing audio and text data from OpenSLR Kashmiri Data Corpus. Includes preprocessing pipelines, data loaders, and segmentation utilities.
Models 3
IndicConformer ASR — Kashmiri
Conformer-Large (120M params) ASR model with hybrid CTC-RNNT decoder specifically for Kashmiri speech. Requires 16kHz mono WAV input. Uses AI4Bharat NeMo framework.
kashurai — Kashmiri Dataset Collection
Curated community collection grouping Kashmiri-specific LLMs, text datasets, and ASR models in one browsable hub. Great starting point for discovery.
XLS-R 300M (Wav2Vec2)
Self-supervised speech model pretrained on 436K hours of audio in 128 languages. Fine-tunable for Kashmiri ASR tasks with limited labeled data.
Papers 5
Task-Lens: Cross-Task Utility Based Speech Dataset Profiling for Low-Resource Indian Languages
Profiles 50 Indian speech datasets in 26 languages for nine speech tasks. Kashmiri has about 170–180 hours across four datasets and is among the languages the authors say urgently need more data (under 400 hours).
A Wav2Vec2 Model-Based Automatic Speech Recognition System for Low-Resource Kashmiri Language
Journal paper on a Wav2Vec2-based speech recognition system for Kashmiri.
Emotion Recognition in Kashmiri Speech Using Bidirectional LSTM Networks
Evaluates coefficient-based acoustic features with bidirectional LSTM networks for recognising emotion in Kashmiri speech — one of the few affective-computing studies targeting the language.
Spoken Kashmiri Recognition Using Hybrid CNN-gMLP
Hybrid CNN-gMLP model for spoken Kashmiri word recognition. Addresses phonetic diversity and dialectal variations with data augmentation strategies.
Kashmiri Speech Recognition Using Linear Predictive Coding and Neural Networks
Early Kashmiri ASR work recognising isolated spoken digits zero to nine across male and female speakers, using LPC feature extraction and artificial neural networks. Reports 92% accuracy on 350 speech samples.
Tools 3
Sarvam Saaras — Kashmiri Speech Recognition
Commercial speech-recognition API from Sarvam AI that lists Kashmiri (ks-IN) among its languages.
Kashmir-ASR
Open-source project for automatic speech recognition of the Kashmiri language, providing code and pipeline scaffolding for building Kashmiri ASR systems.
Bhasini — National Language Translation Mission
Official Indian government AI platform with Kashmiri translation, ASR, and TTS APIs. Free to use for developers. Includes models for all 22 scheduled languages.