Update README.md

README.md

@@ -1,3 +1,26 @@
----
-license: apache-2.0
----
+# This repo is for UTS 42028 Deep Learning and convolutional neural networks Assignment 2 Part A
+
+## Data
+Random seed is set to 24902417
+
+### Classification
+Dataset: [`./dataset/dataset_42028assg2_24902417/Image_Classification`](./dataset/dataset_42028assg2_24902417/Image_Classification)
+No change on the original dataset.
+
+
+### Detection
+For Faster R-CNN, I use the original dataset with different data loading methods.
+For Yolov5, I git clone the Yolov5 repo and convert the original dataset to YOLOv5 format. ([`./yolov5_data_on_coco`](./yolov5_data_on_coco) , [`./pascal_yolov5_format`](./pascal_yolov5_format), [`./yolov5_finetuned_yolo`](./yolov5_finetuned_yolo))
+
+## Model
+
+### Classification
+For classification, All code is in [`classification.ipynb`](classification.ipynb)
+
+### Detection
+For detection, faster-rcnn is in [`./detection.py`](./detection.py)
+yolov5 is in [`./detect_yolo.py`](./detect_yolo.py)
+
+Fast-rcnn training log for different datasets are in [`./training_coco_Faster_R-CNN.log`](./training_coco_Faster_R-CNN.log), [`./training_pascal_Faster_R-CNN.log`](./training_pascal_Faster_R-CNN.log), [`./training_yolo_Faster_R-CNN.log`](./training_yolo_Faster_R-CNN.log)
+
+Yolov5 training log for different datasets are in [`./yolov5_training_results_on_coco`](./yolov5_training_results_on_coco), [`./yolov5_finetuned_yolo`](./yolov5_finetuned_yolo), [`./yolov5_finetuned_pascal`](./yolov5_finetuned_pascal)