Asad

##Hi, Can this code run on video from webcam or IP camera? is there pre-trained network or I should train network then it can be run on video from webcam?...

@KaggleAlbertaAI @HoangTienDuc @12671741 I also have the same problem. How about the speed? is it same?

@ardianumam After several tries, I have successfully optimized this model but the model size and speed is the same. The number of nodes slightly reduces. Can you please check the...

@Diya2507 From the tensorboard these are details about all the model/graph. Which one is the output_names?

@DanielJean007 If you are using YOLOV3 then output nodes are: output_node_names = ["input/input_data", "pred_sbbox/concat_2", "pred_mbbox/concat_2", "pred_lbbox/concat_2"] See the [reference](https://github.com/YunYang1994/tensorflow-yolov3/blob/master/from_darknet_weights_to_pb.py#L8)

> In object detection, we need at least 3 tensors to be called: (i) input, (ii) class prediction and (iii) box prediction. In this case, we need to know those...

@githubljj @yiwenwan2008 I think it's TensorFlow-Yolov3 implementation. Train the model using Tensorflow and then convert .ckpt to .pb. You can use [This project](https://github.com/YunYang1994/tensorflow-yolov3)

+1 , The speed and model size is the same. However, the number of nodes slightly reduced. 

@fugjo16 @ardianumam I have a yolov3 Tensorflow model in both ckpts and .pb format. My model can run in GTX 1080 Ti at 37 FPS . Now I want to...

@johndpope Thank you very much.