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'
Loading...
Searching...
No Matches
Functions | Variables
msecnn_raulkviana.utils Namespace Reference

Functions

 echo (string, padding=80)
 Prints and clears.
 
 update_docstring (path)
 Creates and updates the beginning docstring of python file.
 
 functions_used_from_each_module (str_file)
 Detect the functions associated with each import.
 
 multi_thresholding (rs, preds)
 

Variables

str DOCSTRING_BEGINNING_CREATE
 
str DOCSTRING_BEGINNING_UPDATE
 
str STRING_PATH_1 = "updated_docs"
 
str STRING_PATH_1_2 = "created_docs"
 
str STRING_PATH_2 = ".py"
 
str NONE_STRING = "- None"
 
str AUTHOR = "Raul Kevin Viana"
 

Detailed Description

@package docstring 

@file utils.py 

@brief Usefull functions that are used to not directly for training the model or anything related to this work actual implementation.
 
@section libraries_utils Libraries 
- os
- torch
- datetime
- re

@section classes_utils Classes 
- None
 
@section functions_utils Functions 
- echo(string, padding=80)
- update_docstring(path)
- functions_used_from_each_module(str_file)
- multi_thresholding(rs, preds)
 
@section global_vars_utils Global Variables 
- STRING_PATH_1
- STRING_PATH_1_2
- STRING_PATH_2"
- NONE_STRING
- AUTHOR 

@section todo_utils TODO 
- None 

@section license License 
MIT License 
Copyright (c) 2022 Raul Kevin do Espirito Santo Viana
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

@section author_utils Author(s)
- Created by Raul Kevin Viana
- Last time modified is 2023-01-29 22:22:04.157947

Function Documentation

◆ echo()

msecnn_raulkviana.utils.echo (   string,
  padding = 80 
)

Prints and clears.

Parameters
[in]stringString that will be printed
[in]paddingPadding to avoid concatenations

◆ functions_used_from_each_module()

msecnn_raulkviana.utils.functions_used_from_each_module (   str_file)

Detect the functions associated with each import.

Parameters
[in]str_fileString representation of the file contents
Returns
dict: Dictionary with the following structure - {"'IMPORT #1'": ["'FUNCTION 1#'", "'FUNCTION 2#'", "'FUNCTION 3#'"], "'IMPORT #2'": ["'FUNCTION 1#'", "'FUNCTION 2#'"], ...}

◆ multi_thresholding()

msecnn_raulkviana.utils.multi_thresholding (   rs,
  preds 
)
Accordingly, the multi-threshold values can be chosen in the
following strategies, ensuring the overall prediction accuracy
of MSE-CNN.
• Case 1 (more time saving): if the average threshold
(1/5) * sum(rs) >= 0.4, then τ2 ≥ τ6 ≥ τ3 ≈ τ4 ≈ τ5.
• Case 2 (better RD performance): if the average threshold
(1/5) * sum(rs) < 0.4, then τ2 ≥ τ4 ≈ τ3 ≈ τ5 ≥ τ6.

@brief Implementation of the multi-threshold for MSE-CNN
@param [in] rs: Constant to control the minimum amount of threshold probabilities
@param [in] preds: Predictions made by the MSE-CNN
@param [out] search_RD: Vector with the information of which modes to compute the RD cost and make a decision regarding in which split to make

◆ update_docstring()

msecnn_raulkviana.utils.update_docstring (   path)

Creates and updates the beginning docstring of python file.

Parameters
[in]pathPath to a script

Variable Documentation

◆ AUTHOR

str msecnn_raulkviana.utils.AUTHOR = "Raul Kevin Viana"

◆ DOCSTRING_BEGINNING_CREATE

str msecnn_raulkviana.utils.DOCSTRING_BEGINNING_CREATE

◆ DOCSTRING_BEGINNING_UPDATE

str msecnn_raulkviana.utils.DOCSTRING_BEGINNING_UPDATE

◆ NONE_STRING

str msecnn_raulkviana.utils.NONE_STRING = "- None"

◆ STRING_PATH_1

str msecnn_raulkviana.utils.STRING_PATH_1 = "updated_docs"

◆ STRING_PATH_1_2

str msecnn_raulkviana.utils.STRING_PATH_1_2 = "created_docs"

◆ STRING_PATH_2

str msecnn_raulkviana.utils.STRING_PATH_2 = ".py"