search

myfavouritekk / T-CNN

ImageNet 2015 Object Detection from Video (VID)
MIT License
371 stars 152 forks source link

How to get the tubelet visulaization as good as presented in your ILSVRC 2015 presentation? #3

Closed ronakgupta closed 8 years ago

ronakgupta commented 8 years ago

I ran the demo. After the spatial max-pooling, there are more than one tubelet box appearing on the airplane. I am running the code on Linux platform.

Also, please add commands to run temporal convolution and rescoring

myfavouritekk commented 8 years ago

@ronakgupta Thank you for your feedback. The temporal convolution is not used in this project (but the CVPR paper), and will be included soon in the vdetlib repository.

myfavouritekk commented 8 years ago

@ronakgupta Hi, did you send me an email the other day? Somehow your email address is unknown that causes a delivery failure. Please check your email address. Thanks.

ronakgupta commented 8 years ago

@myfavouritekk Ya, I have again sent the mail. Thanks.

myfavouritekk commented 8 years ago

@ronakgupta still not working, do you have a gmail or something?

ronakgupta commented 8 years ago

@myfavouritekk Ok. Please mail me to ronakgupta143@gmail.com