Japanese Podcasts
Start training, testing or fine-tuning your speech models with 1331 hours of Japanese live, non-simulated podcasts recorded by actual podcasters in our partner network This dataset is perfect for those who are looking for high-quality recordings of spontaneous speech for their ASR or foundational TTS models. Recordings are saved .wav files with a sample rate of 44100 or 48000 and a bit depth of 16 bit. Transcription, either with model or human quality, is available as a service.
Start training, testing or fine-tuning your speech models with 1331 hours of Japanese live, non-simulated podcasts recorded by actual podcasters in our partner network This dataset is perfect for those who are looking for high-quality recordings of spontaneous speech for their ASR or foundational TTS models. Recordings are saved .wav files with a sample rate of 44100 or 48000 and a bit depth of 16 bit. Transcription, either with model or human quality, is available as a service.
Start training, testing or fine-tuning your speech models with 1331 hours of Japanese live, non-simulated podcasts recorded by actual podcasters in our partner network This dataset is perfect for those who are looking for high-quality recordings of spontaneous speech for their ASR or foundational TTS models. Recordings are saved .wav files with a sample rate of 44100 or 48000 and a bit depth of 16 bit. Transcription, either with model or human quality, is available as a service.
Start training, testing or fine-tuning your speech models with 1331 hours of Japanese live, non-simulated podcasts recorded by actual podcasters in our partner network This dataset is perfect for those who are looking for high-quality recordings of spontaneous speech for their ASR or foundational TTS models. Recordings are saved .wav files with a sample rate of 44100 or 48000 and a bit depth of 16 bit. Transcription, either with model or human quality, is available as a service.
Dataset specs
Type
Audio
Sound quality
≥48kHz, 24 bit per channel
Region/Locale
JA,
ja-JP
Amount
171 hours
Leverage
Take your models to the next level. With live, high-quality, Japanese podcast speech data, this dataset is the perfect resource for AI builders working with Conversational AI.
Equip your technologies with the ability to engage in spontaneous dialogue, essential for delivering meaningful interactions to the Japanese-speaking demographic.
Use cases
Train AI models to generate natural-sounding speech from text inputs or to convert written text into spoken audio using the podcast as reference data.
Train LLMs on the podcasts to develop models capable of understanding and generating natural language in the context of natural conversation.
Train AI models to detect emotions and analyze sentiment expressed in the podcast audio.
Do you need a specific dataset? edit
We understand the uniqueness of every project. That's why we offer customizable dataset solutions to match your specific requirements.
Dataset specs
Type
Audio
Sound quality
≥48kHz, 24 bit per channel
Region/Locale
JA,
ja-JP
Amount
171 hours