
Animal Healthcare AI Support
Planning and development support for AI features that classify possible neurological seizures from pet videos.
Technologies Used
GeminiDeepLabCutMachine LearningVideo Processing
Key Features
Pet-video based seizure judgment support
Feature design for seizure classification results
Video embeddings generated with LLMs
Video preprocessing and similarity-method tuning
Overview
This project supported technology and service planning in the animal healthcare domain using AI.
Role
For a pet health management app, Naoki supported planning and development of a feature that uses recorded pet videos to judge whether an event may be an epileptic seizure caused by a neurological disorder, and to return classification results when relevant.
Technical Focus
The system uses LLMs to create embeddings from videos so that they can be used by machine learning models. Work focused on improving accuracy through video preprocessing and similarity-calculation adjustments.
Completed on: May 1, 2025