MSE-CNN Implementation 1
Code database with the implementation of MSE-CNN, from the paper 'DeepQTMT: A Deep Learning Approach for Fast QTMT-based CU Partition of Intra-mode VVC'
|
Dataset stage oriented with capability of loading different files and it's supposed to be used with the function dataset_utils.change_labels_function_again. More...
Public Member Functions | |
__init__ (self, files_path, channel=0) | |
Args: files_path (string): Path to the files with annotations. | |
get_sample (self, entry) | |
Args: | |
![]() | |
__init__ (self, files_path, channel=0) | |
Args: | |
__len__ (self) | |
get_sample (self, entry) | |
Args: | |
obtain_files_sizes (self, files) | |
Args: | |
select_entry (self, idx) | |
Args: | |
__getitem__ (self, idx) | |
Public Attributes | |
channel | |
![]() | |
files_path | |
files | |
lst_entries_nums | |
total_num_entries | |
channel | |
index_lims | |
data_files | |
Dataset stage oriented with capability of loading different files and it's supposed to be used with the function dataset_utils.change_labels_function_again.
For stage 6
This version contains all of the information available from the CU data
msecnn_raulkviana.custom_dataset.CUDatasetStg6Compl.__init__ | ( | self, | |
files_path, | |||
channel = 0 |
|||
) |
Args: files_path (string): Path to the files with annotations.
root_dir (string): Directory with all the images. channel: Channel to get for the dataset (0:Luma, 1:Chroma)
Reimplemented from msecnn_raulkviana.custom_dataset.CUDatasetBase.
msecnn_raulkviana.custom_dataset.CUDatasetStg6Compl.get_sample | ( | self, | |
entry | |||
) |
Args:
entry | (int): An instance from the labels. |
Reimplemented from msecnn_raulkviana.custom_dataset.CUDatasetBase.
msecnn_raulkviana.custom_dataset.CUDatasetStg6Compl.channel |