Update README.md

README.md

@@ -1,14 +1,37 @@
 ---
-title: AMLTestSpace
-emoji: 🚀
+title: Bird Species Classifier
+emoji: 🐦
 colorFrom: blue
-colorTo: blue
+colorTo: green
 sdk: gradio
-sdk_version: 5.49.1
+sdk_version: 4.29.0
 app_file: app.py
 pinned: false
 license: cc-by-2.0
-short_description: For individual assignment, bird classification
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# COS30082 Bird Species Classifier
+
+This project is a deep learning application that classifies bird species from images. It uses a powerful pre-trained model, fine-tuned on the Caltech-UCSD Birds 200 (CUB-200) dataset of 200 different bird species.
+
+## Features
+
+- **High-Accuracy Model**: Over 90% accuracy on the CUB-200 dataset test set.
+- **Multiclass Classification**: The model is able to classify images into one of 200 different bird species, with 99% top-5 accuracy.
+- **Interactive Interface**: Live interface built using Gradio allows users to upload images and get predictions quickly.
+
+## Model Details
+
+- **Model**: `ConvNeXt V2 Large`
+- **Dataset**: Fine-tuned on a private dataset of 200 bird species.
+- **Frameworks**: PyTorch, timm
+- **Notes**: The model is trained with a convoluted process, using model distillation techniques to produce the final model. The approach is likely very unique, and will require extensive training and tuning to replicate or improve, source code available on request.
+
+## How to Use
+
+1.  **Upload an Image**: Click on the "Upload Bird Image" box to select an image file from your computer.
+2.  **Get Predictions**: The model will process the image and display the top 5 most likely bird species, along with their confidence scores.
+
+## License
+
+This project is licensed under the cc-by-2.0 License.