Yet-Another-EfficientDet-Pytorch
Yet-Another-EfficientDet-Pytorch copied to clipboard
zylo117
The pytorch re-implement of the official efficientdet with SOTA performance in real time and pretrained weights.
I am just wondering, why we should initially give model images annotations with JSON format if the model will change it later to VOC. My question is why we need...
Dear @zylo117 , Could you please give us the code which can convert xml annotation to Json(coco) format? I search the code online, and there are several different version. I...
FPS
Hi, First of all, thank for you excellent work. I am using d0 as my pretrained model on my own dataset and got an idea detection outcome on my test-set....
Hi, I have a problem with gpu memory increase I ran your test code(inference) and gpu memory increased 8GB. (when D3) I can't use your bigger model because of this...
@zylo117 Hi! Nice work! 1. Can you explain please, why do you get the result more slowly than in the official paper? 2. Does the official code https://github.com/google/automl/tree/master/efficientdet not reproduce...
I run efficientdet_test.py with d0, but the result is different with your in test folder.,there are a lot of missed objects. And the fps is only 18 with Tesla V100.
Hi, why my infernce speed is so slow? gpu: test in p40 efficientdet-d0 img:512*512 
I have trained my model on a custom dataset. Please let me know if anybody applies the efficientdet_test.py for a trained model on test images. How do you edit this...
Hi Zylo, thanks for providing such a good implementation of Efficientdet. I have a question regarding the activation function in the Separable Convolution, it should be set to true by...
1. fix typo in train.py * admaw -> adamw 2. add multi image test code * can use for multiple images
