Closed SkeletonOne closed 4 years ago
chenjoya
commented
4 years ago Thanks for your attention. That's interesting. It seems that your configuration is incorrect, as training on voc dataset would not call COCO-style evaluation procedure. We have provide "retinanet_voc_R_50_FPN_0.2x.yaml" in configs (configs/retinanet/retinanet_voc_R_50_FPN_0.2x.yaml). Just try it!
SkeletonOne
commented
4 years ago @ChenJoya I used your configs (.yaml) in detectron2 and the annotations were COCO format so the result was like this.... After training pascal voc in its own style I got AP50 about 78.6. Everything seems normal now. Thanks for your work again.
chenjoya
commented
4 years ago Sorry， but it should be 80+ mAP？
------------------ Original ------------------ From: SkeletonOne <notifications@github.com> Date: Mon,Apr 6,2020 9:30 AM To: ChenJoya/sampling-free <sampling-free@noreply.github.com> Cc: ChenJoya <ChenJoya@foxmail.com>, Mention <mention@noreply.github.com> Subject: Re: [ChenJoya/sampling-free] The result of Pascal VOC seems bad. (#6)
@ChenJoya I used your configs (.yaml) in detectron2 and the annotations were COCO format so the result was like this.... After training pascal voc in its own style I got AP50 about 78.6. Everything seems normal now. Thanks for your work again.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.
SkeletonOne
commented
4 years ago Sorry about my unclarity. I was just implementing the original retinanet, without sampling free : )
Since the retinanet paper did not train on Pascal VOC, I was just trying to get the AP of base Retinanet first. Thanks for your kindly reply!
chenjoya
commented
4 years ago Yeah, the vanilla RetinaNet does not seem to do well on PASCAL VOC. Good luck!
------------------ Original ------------------ From: SkeletonOne <notifications@github.com> Date: Mon,Apr 6,2020 4:57 PM To: ChenJoya/sampling-free <sampling-free@noreply.github.com> Cc: ChenJoya <ChenJoya@foxmail.com>, Mention <mention@noreply.github.com> Subject: Re: [ChenJoya/sampling-free] The result of Pascal VOC seems bad. (#6)
Sorry about my unclarity. I was just implementing the original retinanet, without sampling free : )
Since the retinanet paper did not train on Pascal VOC, I was just trying to get the AP of base Retinanet first. Thanks for your kindly reply!
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.
❓ Questions and Help
Hi Joya, Currently, I am training retinanet voc 0.2x as the method you provide in the README. Since the pascal voc data is in COCO format, the default evaluation method is the coco evaluation method. The results are:
Accumulating evaluation results... DONE (t=6.31s). Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.457 Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.722 Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.482 Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.125 Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.321 Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.540 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.398 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.582 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.600 Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.255 Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.510 Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.671 2020-04-01 17:57:02,155 maskrcnn_benchmark.inference INFO: Task: bbox AP, AP50, AP75, APs, APm, APl 0.4572, 0.7220, 0.4823, 0.1246, 0.3206, 0.5403Since the voc result is AP50 in COCO evaluation style, it is only 0.722 here. I am wondering how to get the result you provide. Or in other words, am I possible to use pascal voc evaluation tool in this project or the maskrcnn-benchmark? Maybe that will provide the 79+result. I'm not sure, but how to get the result in your README? Thanks!