grayne

3 ms·

1. 1.▲
   by grayne 9mo ago

   Thanks for the question. audio-to-motion is fairly robust to noisy TTS and differing languages / accents. It doesn't use the raw audio as input, we first embed the audio using a pretrained wav2vec-style embedder, trained on millio
2. 2.▲
   by grayne 9mo ago

   Full technical blog here: https://anam.ai/blog/cara-3-interactive-avatars
3. 3.▲
   Anam Cara-3: Why we think AI needs a face

   24 points by grayne 9mo ago|5 comments
4. 4.▲
   Ask HN: Has anyone been on GCP's springboard programme?

   2 points by grayne 2y ago|0 comments
5. 5.▲
   by grayne 6y ago

   Hi Ben, nice site. I'm no NLP expert but it might be worth taking a look at huggingface; they offer a simple API for all sorts of models. It looks like you'll be interested in their Zero-Shot Classification models in particular, e