Skip to main content
Cross Validated

Return to Question

deleted 33 characters in body; edited tags
Source Link
Ferdi
Ferdi
  • 5.3k
  • 10
  • 47
  • 64

I use the Tensorflow object detection API to train the Pascal VOC dataset from scratch. I just had a look on the first results after 200k training steps and the results are okay, despite that I often have many detections of the same class in Overlapping regions. For example consider the following detections (ignore the wrong person detection in the first image):.

Multiple detection of the same motorcycle Multiple detections of the same aeroplane

Is there a general way to avoid such multiple detecionsdetections of the same object? I guess this is caused by overlapping Region proposals for which the Detection network predicts objects that fit the groundtruth data above the 0.7 IoU threshold, so maybe it would help to set this threshold a bit higher?

Btw. I am using a Faster R-CNN Resnet 101 Architecture.

Edit:

I get a mAP of 0.3 on the whole model. So i think it works quite okay.

I use the Tensorflow object detection API to train the Pascal VOC dataset from scratch. I just had a look on the first results after 200k training steps and the results are okay, despite that I often have many detections of the same class in Overlapping regions. For example consider the following detections (ignore the wrong person detection in the first image):

Multiple detection of the same motorcycle Multiple detections of the same aeroplane

Is there a general way to avoid such multiple detecions of the same object? I guess this is caused by overlapping Region proposals for which the Detection network predicts objects that fit the groundtruth data above the 0.7 IoU threshold, so maybe it would help to set this threshold a bit higher?

Btw. I am using a Faster R-CNN Resnet 101 Architecture.

Edit:

I get a mAP of 0.3 on the whole model. So i think it works quite okay.

I use the Tensorflow object detection API to train the Pascal VOC dataset from scratch. I just had a look on the first results after 200k training steps and the results are okay, despite that I often have many detections of the same class in Overlapping regions. For example consider the following detections (ignore the wrong person detection in the first image.

Multiple detection of the same motorcycle Multiple detections of the same aeroplane

Is there a general way to avoid such multiple detections of the same object? I guess this is caused by overlapping Region proposals for which the Detection network predicts objects that fit the groundtruth data above the 0.7 IoU threshold, so maybe it would help to set this threshold a bit higher?

Btw. I am using a Faster R-CNN Resnet 101 Architecture.

Edit:

I get a mAP of 0.3 on the whole model.

added mAP information.
Source Link
ITiger
ITiger
  • 173
  • 1
  • 8

How does Faster R-CNN handle: How to avoid multiple detection in same area?

I use the Tensorflow object detection API to train the Pascal VOC dataset from scratch. I just had a look on the first results after 200k training steps and the results are okay, despite that I often have many detections of the same class in Overlapping regions. For example consider the following detections (ignore the wrong person detection in the first image):

Multiple detection of the same motorcycle Multiple detections of the same aeroplane

Is there a general way to avoid such multiple detecions of the same object? I guess this is caused by overlapping Region proposals for which the Detection network predicts objects that fit the groundtruth data above the 0.7 IoU threshold, so maybe it would help to set this threshold a bit higher?

Btw. I am using a Faster R-CNN Resnet 101 Architecture.

Edit:

I get a mAP of 0.3 on the whole model. So i think it works quite okay.

How does Faster R-CNN handle multiple detection in same area?

I use the Tensorflow object detection API to train the Pascal VOC dataset from scratch. I just had a look on the first results after 200k training steps and the results are okay, despite that I often have many detections of the same class in Overlapping regions. For example consider the following detections (ignore the wrong person detection in the first image):

Multiple detection of the same motorcycle Multiple detections of the same aeroplane

Is there a general way to avoid such multiple detecions of the same object? I guess this is caused by overlapping Region proposals for which the Detection network predicts objects that fit the groundtruth data above the 0.7 IoU threshold, so maybe it would help to set this threshold a bit higher?

Btw. I am using a Faster R-CNN Resnet 101 Architecture.

Faster R-CNN: How to avoid multiple detection in same area?

I use the Tensorflow object detection API to train the Pascal VOC dataset from scratch. I just had a look on the first results after 200k training steps and the results are okay, despite that I often have many detections of the same class in Overlapping regions. For example consider the following detections (ignore the wrong person detection in the first image):

Multiple detection of the same motorcycle Multiple detections of the same aeroplane

Is there a general way to avoid such multiple detecions of the same object? I guess this is caused by overlapping Region proposals for which the Detection network predicts objects that fit the groundtruth data above the 0.7 IoU threshold, so maybe it would help to set this threshold a bit higher?

Btw. I am using a Faster R-CNN Resnet 101 Architecture.

Edit:

I get a mAP of 0.3 on the whole model. So i think it works quite okay.

Source Link
ITiger
ITiger
  • 173
  • 1
  • 8

How does Faster R-CNN handle multiple detection in same area?

I use the Tensorflow object detection API to train the Pascal VOC dataset from scratch. I just had a look on the first results after 200k training steps and the results are okay, despite that I often have many detections of the same class in Overlapping regions. For example consider the following detections (ignore the wrong person detection in the first image):

Multiple detection of the same motorcycle Multiple detections of the same aeroplane

Is there a general way to avoid such multiple detecions of the same object? I guess this is caused by overlapping Region proposals for which the Detection network predicts objects that fit the groundtruth data above the 0.7 IoU threshold, so maybe it would help to set this threshold a bit higher?

Btw. I am using a Faster R-CNN Resnet 101 Architecture.