Assistants: 

Siri, Alexa, Google Assistant

Search: Google, Bing suggestions

Recommendations: Netflix, Spotify

Navigation: Google Maps

Social Media: Personalized feeds

Shopping: Product suggestions

Customer Support: Chatbots

Banking: Fraud detection

Smart Homes: Devices like smart thermostats

Language: Google Translate

1. Text-to-Text

-Models take text as input and generate text as output

-You enter the prompt in text, then you get your result in text

-Examples: Chatbot

2. Text-to-Image

-Models take text as input and generate image as output

-You enter the prompt in text, then you get your result as images

-DALL·E: Generates images from detailed text prompts.

-Examples: AI painting

3. Text-to-Speech

-Models that convert text into audio 

-Some tools converts text into natural-sounding audio, such as Google Text-to-Speech

4. Text-to-Video

-Models take text as input and generate videos as output

-You enter the prompt in text, then you get your result in videos

-Examples: Sora

5. Image-to-Text

-Models that generate textual descriptions from images

-Examples: Captioning models

6. Image-to-Image

-Models that transform one image into another.

-Uses: Style transfer, image restoration, image upscaling

7. Image-to-3D

-Models that convert 2D images into 3D models

-Examples: NVIDIA Omniverse–Generates 3D scenes from 2D sketches or images.

 8. Speech-to-Text

-Models that transcribe spoken language into written text.

-Examples: Google Speech-to-Text: Converts audio to text.

9. Speech-to-Speech

-Models take audio as input and generate another track of audio as output

-Examples: Voice cloning–Replicating a speaker’s voice, real-time translation, voice assistance

10. Speech-to-Image

-Models that generate images based on spoken descriptions

-Combining text-to-image AI with speech-to-text capabilities

11. Video-to-Text

-Models that analyze video content and generate textual descriptions or summaries.

-Examples: Video Captioning–Generating descriptions for video scenes, Action Recognition–Identifying activities in videos

12. Video-to-Video

-Models that enhance, modify, or create new videos from existing ones

-Examples: Deepfakes–Replacing faces in videos, Super-resolution for Video–Improving video quality.