This is a submission for Weekend Challenge: Dog Days Edition
What I Built
Translate My Dog is a multimodal AI web application that reads your dog's subtle facial micro-expressions, ear positions, and body posture from photos to reveal what they are actually thinking. It generates witty, character-accurate inner monologues and synthesizes character voice audio.
Demo
Code
maneesh-kumar-thakur
/
devto-challenge-dog-days-edition
dev.to weekend challenge
🐾 Translate My Dog (AI Canine Micro-Expression Reader)
Translate My Dog is a multimodal AI web application that reads your dog's subtle facial micro-expressions, ear positions, and body posture from photos to reveal what they are actually thinking. It generates witty, character-accurate inner monologues and synthesizes character voice audio.
🌟 Key Features
1. 🔍 Multimodal Canine Micro-Expression Vision Analysis
- Powered by Google Gemini Vision models (
gemini-2.5-flash,gemini-flash-lite-latest, etc.). - Evaluates canine facial cues, pupil dilation, ear angles, head tilts, tail dynamics, and environmental clues (e.g. vacuum cleaners, treat jars, tennis balls).
- Outputs a comprehensive translation report
- First-Person Inner Monologue: Spoken directly in character as the dog.
- Detected Mood Title: Humorous real-time emotional state (e.g., "94% Betrayed By Diet Kibble").
- Identified Breed & Heritage Insight: Visual breed detection with evolutionary and behavioral notes.
- Visual Clues Breakdown: 3 specific visual observations justifying the monologue.
- Canine…
How I Built It
Prize Categories
Applying for Best Use of ElevenLabs & Best Use of Google AI
Top comments (0)