print1Username commited on
Commit
6ff65ad
·
verified ·
1 Parent(s): 5b7f27c

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +42 -1
README.md CHANGED
@@ -3,7 +3,48 @@ datasets:
3
  - print1Username/COS30082-Assignment-1
4
  language:
5
  - en
 
6
  base_model:
7
  - microsoft/resnet-50
8
  pipeline_tag: image-classification
9
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3
  - print1Username/COS30082-Assignment-1
4
  language:
5
  - en
6
+ - zh
7
  base_model:
8
  - microsoft/resnet-50
9
  pipeline_tag: image-classification
10
+ tags:
11
+ - image-classification
12
+ - pytorch
13
+ metrics:
14
+ - accuracy
15
+ library_name: timm
16
+ ---
17
+
18
+ # COS30082 Assignment 1 - Bird Species Classification
19
+
20
+ ## Model Description
21
+ This is an image classification model trained to identify 200 different bird species from the Caltech-UCSD Birds 200 (CUB-200) dataset. This model was developed as part of the COS30082 Assignment 1.
22
+
23
+ ## Intended Use & Limitations
24
+ - **Intended Use**: Academic purpose for predicting bird species based on images.
25
+ - **Limitations**: The model is specifically trained on North American bird species included in the CUB-200 dataset. It may not perform well on bird species outside this dataset or in significantly different environments/lighting conditions.
26
+
27
+ ## Training Data
28
+ The model was trained using the **Caltech-UCSD Birds 200 (CUB-200)** dataset, which contains photos of 200 bird species.
29
+ - Dataset source: http://www.vision.caltech.edu/visipedia/CUB-200.html
30
+
31
+ ## Evaluation Results
32
+ Based on the testing dataset provided, the model achieves the following performance metrics as required by the assignment guidelines:
33
+ - **Top-1 accuracy**: 78.5%
34
+ - **Average accuracy per class**: 77.2%
35
+
36
+ ## How to Use
37
+ Here is how to load and use the model in PyTorch:
38
+
39
+ ```python
40
+ import torch
41
+ from huggingface_hub import hf_hub_download
42
+
43
+ # Define your model architecture first
44
+ # model = MyBirdClassifier()
45
+
46
+ # Download and load the weights
47
+ model_path = hf_hub_download(repo_id="print1Username/COS30082-Assignment-1-model", filename="bird_model.pth")
48
+ model.load_state_dict(torch.load(model_path, map_location="cpu"))
49
+ model.eval()
50
+ ```