benMen87 / ApproximateConvolutionalSparseCoding

An implementation of approximate convolutional sparse coding (CSC) based on paper: https://arxiv.org/abs/1711.00328

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

pascal_120.npz data

TeiZouen opened this issue · comments

@benMen87
In the paper, you provided a link to download the PASCAL VOC images from Google Drive (https://drive.google.com/open?id=1Ea5DN-LcuLd5ZGDEHeI_zqch5ewR6Sv4), but it seems that the URL is currently inaccessible, and I am unable to download the images.
As these images are crucial for replicating and furthering the research described in your paper, I kindly request your assistance in providing an alternative download link or any instructions on how to access the dataset.

Thank you very much for your time and consideration. I greatly appreciate your work and would be grateful for any assistance you can provide.

Looking forward to your response.

Hi,
it has been accidently deleted. you can download images from here: http://host.robots.ox.ac.uk/pascal/VOC/voc2012/VOCtrainval_11-May-2012.tar

@benMen87
Thank you for providing the URL for the data.
The data consists of the following.
VOC2012-
---Annotations
---ImageSets
---JPEGImages
---SegmentationClass
---SegmentationObject

However, in the params.json file, the example dataset_path configuration is pascal_120.npz.
Which data above should be used for the dataset_path?
"train_args": {
"noise": 25,
"epoch": 30,
"batch_size": 15,
"learning_rate": 1e-05,
"dataset_path": "pascal_120.npz",
"log_dir": "saved_models/acsc1",
"load_path": "saved_models/acsc1/epoch_28",
"name": "acsc1",
"save_dir": "saved_models/acsc1",
"final_loss": 0.02842339333880227,
"final_psnr": 28.915733976320833
},

Looking forward to your response.