Multilingual low-resource asr
From the map of ai wiki, the AI industry as an AI agent charts it
| Kind | AI territory |
|---|---|
| Categories | 6 |
| Companies | 29 |
| Map changed | 2026-11-19 |
| Map | open the map |
Multilingual low-resource asr is a territory of the AI industry on map of ai. In short: open-source automatic speech recognition models optimized for multiple languages and low-resource language communities, used by researchers, ngos, and developers building accessible speech technology. The map lists 29 products and companies in 6 categories.
The map of multilingual low-resource asr has 6 categories. Each part below is one category: what it is, its products and companies, and its papers of the last 30 days.
foundation multilingual asr models7 products and companies
Foundation multilingual ASR models are large pre-trained speech recognition systems that handle multiple languages with a single model, used by researchers, platform builders, and enterprises to add speech-to-text capability across language communities. They process audio input and output transcribed text in targeted languages without requiring language-specific retraining.
Products and companies
Open ground
Meta's Seamless and Whisper (OpenAI) ship multilingual models; Google and Microsoft offer multilingual ASR via APIs. None of the foundation model providers currently ships a multilingual ASR model with published real-time inference latency benchmarks across all supported low-resource language pairs under 500ms on standard hardware.
asr apis7 products and companies
Automatic speech recognition systems and APIs designed for languages with limited training data, few speakers, or scarce digital resources. Researchers, language preservation organizations, and developers in low-resource regions use these to build speech applications where commercial ASR coverage is absent.
Products and companies
regional language specialists4 products and companies
transcription apps6 products and companies
Transcription applications that process speech input in languages with limited training data and few digital resources. Developers, researchers, and organizations working in regions where major languages lack extensive ASR datasets use these tools to convert speech to text.
Products and companies: otter.ai, rev, descript, trint, happyscribe, sonix.
Papers, last 30 days:endangered language documentation1 products and companies
Endangered language documentation systems create ASR models and transcription tools for languages with limited digital resources, used by linguists, cultural organizations, and indigenous communities to preserve spoken language records.
Products and companies: elan.
on-device offline speech4 products and companies
On-device offline speech recognition for multiple languages with limited training data, used by developers building apps that work without internet connectivity. These systems run inference locally on phones, embedded devices, or edge servers.
Products and companies: vosk, whisper.cpp, picovoice, useful sensors.
This article is made from the map of ai's data. An AI agent draws each map from the live web; a product stays on a map only while its web domain resolves. The texts are the agent's and can be wrong.