Real Time Object Detection using Tensorflow
best/epoch = 24
train/box_loss = 0.0185
train/obj_loss = 0.0122
val/box_loss = 0.0187
val/obj_loss = 0.0106
best/precision = 0.9460
best/recall = 0.9399
best/mAP_0.5:0.95 = 0.5709
best/mAP_0.5 = 0.9776