search

guichristmann / edge-tpu-tiny-yolo

Run Tiny YOLO-v3 on Google's Edge TPU USB Accelerator.
MIT License
102 stars 31 forks source link

※important※ Mistake in iou() of util.py #21

Open YsYusaito opened 3 years ago

YsYusaito commented 3 years ago
I added exception handling. ``` def iou(box1, box2): xi1 = max(box1[0][0], box2[0][0]) yi1 = max(box1[0][1], box2[0][1]) xi2 = min(box1[1][0], box2[1][0]) yi2 = min(box1[1][1], box2[1][1]) if(((xi2 - xi1) < 0) or ((yi2 - yi1) < 0)): IoU = 0 return IoU else: inter_area = (xi2 - xi1)*(yi2 - yi1) # Formula: Union(A,B) = A + B - Inter(A,B) box1_area = (box1[1][1] - box1[0][1])*(box1[1][0]- box1[0][0]) box2_area = (box2[1][1] - box2[0][1])*(box2[1][0]- box2[0][0]) union_area = (box1_area + box2_area) - inter_area # compute the IoU IoU = inter_area / union_area return IoU ```
AkkiSony commented 3 years ago

@YsYusaito Hi. Did you get yolov3 running using inference script? I could not get any object detection suing the inference script. Please let me know if you got it running. Thanks in advance. :)

YsYusaito commented 3 years ago

@AkkiSony Hi, I executed yolov3 object detection using inference script. For reference, I wrote my inference environment in issue #23.

AkkiSony commented 3 years ago

@YsYusaito Thank you very much for sharing your environment issue. Did you make any changes in the inference.py and utils.py script for getting the objects detected. Please let me know if so. Would it be possible for you to share your utils.py and inference.py script? It would really help me to get yolov3 run on coral TPU. :) Thanks in advance! :)