Very explanatory class. Thank you! Just a subtle contribution: in the video, the mAP is calculated considering a single threshold (lets say, 0.5), so it would be mAP@0.5. In papers it's common to see also a metric called mAP@0.5:0.05:0.95, which is the average of the calculations of mAP@0.5, mAP@0.55, mAP@0.60, ... , mAP@0.95. One more time thank you for this awesome content!
What is the method of selecting these red bounding boxes ? Is there any step between selective search which suggested 2000 boxes and counting mean average precision ?As i guess IOU is counted after suggestion of bounding boxes .Which algorithm is responsible for confidence scores,how it assessed ? Thanks
I think it's one of the best explanations of mAP which I have seen on the internet. thanks a lot!
Very explanatory class. Thank you!
Just a subtle contribution: in the video, the mAP is calculated considering a single threshold (lets say, 0.5), so it would be mAP@0.5. In papers it's common to see also a metric called mAP@0.5:0.05:0.95, which is the average of the calculations of mAP@0.5, mAP@0.55, mAP@0.60, ... , mAP@0.95.
One more time thank you for this awesome content!
Great work!
What is the method of selecting these red bounding boxes ? Is there any step between selective search which suggested 2000 boxes and counting mean average precision ?As i guess IOU is counted after suggestion of bounding boxes .Which algorithm is responsible for confidence scores,how it assessed ? Thanks
Thank you