https://github.com/saptakbhoumik/tinyvision

TinyVision is an evolving project focused on designing ultra-lightweight image classification models with minimal parameter counts. The goal is to explore what’s actually necessary for fundamental vision tasks by combining handcrafted feature preprocessing with highly efficient CNN architectures.
https://github.com/saptakbhoumik/tinyvision

computer-vision machine-learning python python3 pytorch vision

Last synced: 11 months ago
JSON representation

TinyVision is an evolving project focused on designing ultra-lightweight image classification models with minimal parameter counts. The goal is to explore what’s actually necessary for fundamental vision tasks by combining handcrafted feature preprocessing with highly efficient CNN architectures.

• Host: GitHub
• URL: https://github.com/saptakbhoumik/tinyvision
• Owner: SaptakBhoumik
• License: cc-by-sa-4.0
• Created: 2025-07-27T00:16:10.000Z (11 months ago)
• Default Branch: master
• Last Pushed: 2025-08-06T15:45:29.000Z (11 months ago)
• Last Synced: 2025-08-06T17:44:42.719Z (11 months ago)
• Topics: computer-vision, machine-learning, python, python3, pytorch, vision
• Language: Jupyter Notebook
• Homepage:
• Size: 10.5 MB
• Stars: 10
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# 🧠 TinyVision: Compact Vision Models with Minimal Parameters

**TinyVision** is an evolving research project focused on designing **ultra-lightweight image classification models** with minimal parameter counts. The goal is to explore what’s *actually necessary* for fundamental vision tasks by combining **handcrafted feature preprocessing** with highly efficient CNN architectures.

📦 **Current Release**: [v2.0.0](https://github.com/SaptakBhoumik/TinyVision/releases/tag/v2.0.0) 

🔖 **Zenodo DOI**: [10.5281/zenodo.16467349](https://doi.org/10.5281/zenodo.16467349)  

📁 **Latest Results & Code**: See the `cat_vs_dog_classifier/final/v2` directory

> ⚠️ This release **does not include a paper**, but focuses on the **codebase**, experiment results, and reproducible training scripts. A deeper analysis and formal documentation will come in future updates.

---

## 🚧 Project Status

- ✅ **Cat vs Dog Classification**  

  First completed task using a 25,000-image dataset with handcrafted preprocessing + compact CNNs.  

  - Achieved **up to 86.87% test accuracy** with models under **12.5k parameters**

  - Several models under **5k parameters** reached over **83% accuracy**, showcasing strong efficiency-performance trade-offs.

  - 📂 Final results and code for this task are in the `cat_vs_dog_classifier/final/v2` directory.

---

## 🧪 What's Coming Next

- 📊 Add thorough **performance analysis** of model architectures to understand why something works while others don't

- 🧩 Explore new **vision tasks** (edge detection, object detection, etc.) with compact models

- 📖 Expand **documentation**, architecture diagrams, and visualizations

- 🧠 Log and reflect on **failed or inconclusive experiments** critical for understanding design boundaries

---

## 🤝 Contributing

This project is currently personal and tracks my ongoing experiments.  

I’m **not accepting pull requests**, but you're welcome to:

- 📬 Open an [issue](https://github.com/SaptakBhoumik/TinyVision/issues) for discussion or feedback  

- 💌 Reach me at: `saptakbhoumik.acad@gmail.com`

- 📢 Follow me on [X](https://x.com/saptakbhoumik)

---

## 💡 Philosophy

> Small models aren't just about speed—they’re a design challenge.  

> *How much can we cut before it breaks? What’s essential? What’s fluff?*

TinyVision is my attempt to find those answers.

---