测试记录

轻量化神经网络测试

尝试训练100张640x480大小的图片

网络模型	Time(s)	Images/s
Openpose(VGG-19)	36.4	2.75
Openpose(MobileNet v1)	4.2	23.94

尝试训练100张224x224大小的图片

网络模型	Time(s)	Images/s
Openpose(VGG-19)	10.45	9.57
Openpose(MobileNet v1)	1.9	53.68

在224x224分辨率图像输入下网络的参数量和运算量

网络模型	参数量(Million)	GFLOPs
Openpose(VGG-19)	52.31	50.41
Openpose(MobileNet v1)	2.88	50.41

使用验证集(Val )进行验证

 Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets= 20 ] = 0.400
 Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets= 20 ] = 0.660
 Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets= 20 ] = 0.407
 Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets= 20 ] = 0.338
 Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets= 20 ] = 0.494
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 20 ] = 0.462
 Average Recall     (AR) @[ IoU=0.50      | area=   all | maxDets= 20 ] = 0.698
 Average Recall     (AR) @[ IoU=0.75      | area=   all | maxDets= 20 ] = 0.476
 Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets= 20 ] = 0.359
 Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets= 20 ] = 0.605