the validation app provides for Object Detection the Mean Average Precision. I can not find any information in Deep Learning literature about the Max Integral metric for Object Detection, that I could report to our dev team for a future feature. Can you send me such info (link to papers, etc...)?
Thanks for the reply! I found the implementation of MaxIntegral in Wei Liu's SSD repo:
You could also see this link for a more detailed explanation:
Hope that can help!