Kevin Ke-Yun Lin

Could you please describe more details about your system setup and the AP scores you got using —eval_method 0? In addition, could you please post your logs with --eval_method 1?...

Great. Thanks for the logs. One possible reason is that the eval codes doesn't load the config file correctly, and it doesn't setup the multi-scale range (which should be scale_search...

I am wondering that did you conduct performance evaluation on the public datasets such as COCO keypoint dataset? I feel this could be the a best way to validate the...

@anatolix Thank you for this awesome finding. I followed your suggestions and trained the model on COCO train2014 dataset. I strictly followed COCO eval protocol and successfully reproduced similar performance...

From my understanding, center-cropping each person removes the background noise, some of the occlusions, and helps the network concentrate more on learning the keypoints for each person. Please correct me...

@GaogaoT You may want to rebuild your coco api using python3.x 1. go to `coco/PythonAPI` 2. Setup your coco api using `sudo python3 setup.py build_ext --inplace` under `coco/PythonAPI` 3. Then,...

The human detector is not integrated in the code base.

Thanks for your interest in our previous work. You may want to cite the following paper: ``` Abandoned Object Detection via Temporal Consistency Modeling and Back-Tracing Verification for Visual Surveillance...

Which OpenCV are you using? I am using OpenCV-2.3.1

Could you downgrade OpenCV and compile again?