CoolFace
Apppublic

ainulyaqinmhd/SpatialReasoning

sourceHugging Faceupdated 1y agoView on Hugging Face
0likes
App README

🧠 Multimodal AI Agent: Image + Text + Voice → Text + Natural Voice

This Hugging Face Space accepts:

  • —Uploaded Image
  • —Text Prompt
  • —Uploaded Audio file (voice)

It then generates:

  • —AI-written Text output
  • —Human-like Voice Output (using Hugging Face Text-to-Speech model)

Models Used

  • —Image-to-Text: Salesforce/blip-image-captioning-base
  • —Speech-to-Text: openai/whisper-small
  • —Text Generation: gpt2
  • —Text-to-Speech: espnet/kan-bayashi_ljspeech_vits

How to Use

  1. 1.Upload an image.
  2. 2.Enter a text prompt.
  3. 3.Upload a voice file.
  4. 4.Click "Generate".
  5. 5.Receive AI-generated text and voice audio.

Created with ❤️ using Gradio + Hugging Face Transformers.