TDY-PKG 1.1.4

its an implimentation of TF-2 , Detectron and yolov5

Package for TF2 , Detectron2 and Yolov5 Inferencing.

This packge can be useful to do inferencing on TF2, Detectron2 and yolov5. All the supported libraries are already added in the package.
User can select five pretrained model from each framework to do prediction.

TF2

1. ssd_mobilenet_v1_fpn_640x640_coco17_tpu-8
   
2. ssd_resnet101_v1_fpn_1024x1024_coco17_tpu-8
   
3. efficientdet_d4_coco17_tpu-32
   
4. faster_rcnn_resnet101_v1_1024x1024_coco17_tpu-8
   
5. efficientdet_d7_coco17_tpu-32
   

Detectron2

1. faster_rcnn_R_50_FPN_1x
   
2. faster_rcnn_R_50_FPN_3x
   
3. faster_rcnn_R_101_FPN_3x
   
4. retinanet_R_50_FPN_1x
   
5. faster_rcnn_R_50_C4_1x
   

Yolov5

1. yolov5n
2. yolov5s
3. yolov5m
4. yolov5l
5. yolov5x

How to Run?

Step 1:

pip install TDY-PKG

Step 2: Copy the templates, utils folder and app.py file in any directory from the Github link

Step 3: Run app.py file.

python app.py

Step 4: Go to : http://127.0.0.1:9502/

Step 5: Upload Image and select Framework and Model.

Step 6: Click On Predict.(The model will get download in runtime to do prediction)