The goal of this project was to train an image recognition model using my personal collection of wildlife photographs. This approach differs from most image recognition models, which are typically trained on thousands of photos gathered from various sources. By using my own photos, the model becomes a personalized image recognition tool, offering a glimpse into my life and photography . This project is a fusion of my three passions: photography, machine learning, and programming.
The model is designed to recognize my personal animals rather than some animal. For instance:
- It recognises the neighbor's cat/the korean neighbor's cat rather than thinking it's just some cat
- It recognises Pomegranate, my sister's cat, whom should be categorically distinct from neighbor's cat
- It recognises Naruto, Sasuke, and Co, my fishes, they are not food, they are more than just fish, and the image recognition agrees with me.
TLDR: This project intersects my interests, my programming knowledge, and personal life. Also I realised Pomegranate and Co are technically not wildlife.
Pomegranate ↓
/20230521_191927.jpg)