Dhrumit1314 commited on
Commit
7e15528
1 Parent(s): 43af232

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +32 -0
README.md CHANGED
@@ -1,3 +1,35 @@
1
  ---
2
  license: mit
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: mit
3
  ---
4
+ # Food Vision with EfficientNet
5
+
6
+ This repository contains the code for the Food Vision project using EfficientNet. The project involves building a deep learning model to classify food images into 101 different classes using the Food101 dataset.
7
+
8
+
9
+ ## Project Overview
10
+
11
+ This project utilizes TensorFlow and EfficientNet for image classification. It involves training a model on the Food101 dataset, fine-tuning the model, and evaluating its performance.
12
+
13
+ ## Dataset
14
+
15
+ The [Food101 dataset](https://www.tensorflow.org/datasets/catalog/food101) is used for this project. It consists of 101,000 images across 101 food classes.
16
+
17
+ ## Data Preprocessing
18
+
19
+ The dataset is preprocessed using TensorFlow Datasets (TFDS). Images are resized, normalized, and batched to create an efficient input pipeline for the model.
20
+
21
+ ## Model Architecture
22
+
23
+ The EfficientNetV2B0 architecture is used as the base model for feature extraction. The top layers are added for classification. The model is compiled with a suitable loss function, optimizer, and metrics.
24
+
25
+ ## Training
26
+
27
+ The model is trained on the preprocessed data, and the training process is logged using TensorBoard. Checkpoints are saved to monitor the model's progress.
28
+
29
+ ## Fine-tuning
30
+
31
+ After feature extraction, the model is fine-tuned on the entire Food101 dataset. Learning rate reduction and early stopping callbacks are used to optimize training.
32
+
33
+ ## Results
34
+
35
+ The model's performance is evaluated on the test set, and the results are compared before and after fine-tuning.