Abhijeet Pise

icon company @t-systems

Results 26 comments of Abhijeet Pise

the use of Lambda layer makes saving model and weights useless, how to resovle this?

Hi @WangRuixin93 Please use TF 1.15 and Keras 2.1.5.

the full data website can't be opened

Yes, same issue here. Is there a solution?

Using Blender Script to Make Datasets

> Does someone tell me how to use these blender file？ Download Blender 2.92 and open the .blend script. Change the default ModelNet paths and modify the code if you...

Where can I find yolov3_intersting_layers.txt?

@Sibozhu Quite useful repo. Could you please upload the same?

Where can I find yolov3_intersting_layers.txt?

@Sibozhu Could you also make the parameters dynamic? So as to we could use any .weights and .cfg file from a custom detector? Thanks

The way `--output_coco_json` works right now is it operates on a dataset (which doesn't need to have ground truth). So you need to give a `--dataset` parameter instead of an `--image` parameter. The dataset you use doesn't need annotations, simply create a COCO dataset with just images and an empty annotation array and follow the dataset definition of `coco2017_testdev_dataset`.

Hi @dbolya , Same here, is there a way to get a single json file (COCO format) per image as output of the inference? Thank you :)

The way `--output_coco_json` works right now is it operates on a dataset (which doesn't need to have ground truth). So you need to give a `--dataset` parameter instead of an `--image` parameter. The dataset you use doesn't need annotations, simply create a COCO dataset with just images and an empty annotation array and follow the dataset definition of `coco2017_testdev_dataset`.

Hi @PL-96, @linyingyingkarina I could use the dataset parameter as suggested and it gave me two files: mask_detections.json bbox_detections.json How can i get single file per image from these two...

The way `--output_coco_json` works right now is it operates on a dataset (which doesn't need to have ground truth). So you need to give a `--dataset` parameter instead of an `--image` parameter. The dataset you use doesn't need annotations, simply create a COCO dataset with just images and an empty annotation array and follow the dataset definition of `coco2017_testdev_dataset`.

Here is my eval command: python eval.py --trained_model=weights/yolact_base_540_20000.pth --score_threshold=0.6 --top_k=50 --dataset=my_custom_dataset --output_coco_json

The way `--output_coco_json` works right now is it operates on a dataset (which doesn't need to have ground truth). So you need to give a `--dataset` parameter instead of an `--image` parameter. The dataset you use doesn't need annotations, simply create a COCO dataset with just images and an empty annotation array and follow the dataset definition of `coco2017_testdev_dataset`.

Hi @PL-96, I am getting the segmentation in encoded format: ![image](https://user-images.githubusercontent.com/57705684/115408454-7e787f00-a20e-11eb-9d65-4797711ad83a.png) Is there a way to deal with it?

The way `--output_coco_json` works right now is it operates on a dataset (which doesn't need to have ground truth). So you need to give a `--dataset` parameter instead of an `--image` parameter. The dataset you use doesn't need annotations, simply create a COCO dataset with just images and an empty annotation array and follow the dataset definition of `coco2017_testdev_dataset`.

Hi @PL-96 No, here is what i want to do: Manual annotation - drawing polygons is a painful task. I have trained the YOLACT model on few images and it...