Zhangxuan Gu

Train CenterNet with this code

1

Hi! It's a nice work!! I wonder if I can train a CenterNet for image detection on coco? I train the model with ''' python main.py ctdet --exp_id coco_dla_1x --batch_size...

ValueError: Shape must be rank 2 but is rank 3 for 'MatMul' (op: 'MatMul') with input shapes: [?,1000,2048], [2048,64].

1

The error raised when I run this code: ''' import loupe as lp import tensorflow as tf x = tf.placeholder("float", [None,1000,2048]) NetVLAD = lp.NetVLAD(feature_size=2048, max_samples=1000, cluster_size=64, output_dim=2048, gating=True, add_batch_norm=True, is_training=True)...

What is the FPS of this model during inference based on Res-50? Thx!!

Any other results on other datasets like NYUDv2?

1

I find some dataloaders in this code and I wonder have you ever done some experiments on these datasets? Like NYUDv2? Is your method still work on these datasets? Thanks...

Hi, thanks for your nice work! I want to know how to use annotated points to get instance mask? I can not find this code in the git..