search

matterport / Mask_RCNN

Mask R-CNN for object detection and instance segmentation on Keras and TensorFlow
Other
24.67k stars 11.71k forks source link

Finding accuracy and plotting the learning curve and loss graph in mask rcnn #2826

Open filmonmehari opened 2 years ago

filmonmehari commented 2 years ago

I am just using maskrcnn model for my project. But i am unable to find out the accuracy of my model and plot the learning curve( train and validation Vs epoc) and loss graph (i.e train loss vs test loss). I have read previous discussions abt accuracy saying mAP is used in object detection, but my advisor is requesting me to show him the accuracy(he is familiar only with the normal accuracy metrics) , learning curve and loss graph.

Even i am troubled whether i have to add data for my model cuz i didn't find the top accuracy and learning curve in mask rcnn. So please help me for the solution , i have only 3 weeks to submit with those parameters.

filmonmehari commented 2 years ago

Thanks in advance!!!!!

filmonmehari commented 2 years ago

Moreover I'm watching the losses on the output but how to see those using a graph after training for analysis. I have seen this is possible in one journal "Use of Very High Spatial Resolution Commercial Satellite Imagery and Deep Learning to Automatically Map Ice-Wedge Polygons across Tundra Vegetation Types" but i want to know the code for that.