A CNN-based QTMT partitioning decision for the VVC standard

Abdallah, Bouthaina; Jdidia, Sonda Ben; Belghith, Fatma; Ayed, Mohamed Ali Ben; Nouri, Mohammad

doi:10.1109/dts55284.2022.9809888

Cited by 3 publications

(3 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Tissier et al [ 29 ] divided a 64 × 64 CU into 256 4 × 4 CUs and used CNN to predict the probability that the boundaries of each 4 × 4 CU existed. Abdallah et al [ 30 ] proposed a fast CU partition algorithm based on CNN and designed a neural network structure named CNN-BTH to predict the decision depth of 32 × 32 CU horizontal binary tree division. Chen et al [ 31 ] used the pixel variance value of the original image to terminate the further division of the CU with a size of 32 × 32 in advance.…”

Section: Background and Related Workmentioning

confidence: 99%

A Fast Decision Algorithm for VVC Intra-Coding Based on Texture Feature and Machine Learning

Zhao

Zhang

2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

Due to the development and application of information technology, a series of modern information technologies represented by 5G, big data, and artificial intelligence are changing rapidly, and people’s requirements for video coding standards have become higher. In the High-Efficiency Video Coding (HEVC) standard, the coding block division is not flexible enough, and the prediction mode is not detailed enough. A new generation of Versatile Video Coding (VVC) standards was born. VVC inherits the hybrid coding framework adopted by HEVC, improves the original technology of each module, introduces a series of new coding technologies, and builds on this greatly improving the coding efficiency. Compared with HEVC, the block division structure of VVC has undergone great changes, retaining the quad-tree (QT) division method and increasing the multi-type tree (MTT) division method, which brings high coding complexity. To reduce the computational complexity of VVC coding block division, a fast decision algorithm for VVC intra-frame coding based on texture characteristics and machine learning is proposed. First, we analyze the characteristics of the CU partition structure decision and then use the texture complexity of the CU partition structure decision to terminate the CU partition process early; for CUs that do not meet the early termination of the partition, use the global sample information, local sample information, and context information. The three-category feature-trained tandem classifier framework predicts the division type of CU. The experimental results show that in the full intra mode, compared with the existing VTM10.0, the encoding output bit rate is increased by 1.36%, and the encoding time is saved by 52.63%.

show abstract

Section: Background and Related Workmentioning

confidence: 99%

A Fast Decision Algorithm for VVC Intra-Coding Based on Texture Feature and Machine Learning

Zhao

Zhang

2022

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

show abstract

“…For example, the QTMT block division structure and the number of intra-frame prediction modes are increased to 67, which not only improves the coding efficiency, but also increases the computational complexity of the coding process. As described in [2], the coding complexity of VVC is about 18 times higher than that of HEVC when intra coding. The official team named VVC's reference software Versatile Video Coding Test Model (VTM).…”

Section: Introductionmentioning

confidence: 99%

“…Our contributions are summarized as follows: (1) We use an improved DenseNet model, which can make full use of the global features of CU and further improve the accuracy of extracting spatial feature vectors. (2) We utilize the state-of-the-art LGBM classifier for classifier selection. LGBM demonstrates exceptional capability in processing large-scale data.…”

Section: Introductionmentioning

confidence: 99%

Fast CU Decision Algorithm Based on CNN and Decision Trees for VVC

Zhang

Jin

et al. 2023

Electronics

View full text Add to dashboard Cite

Compared with the previous generation of High Efficiency Video Coding (HEVC), Versatile Video Coding (VVC) introduces a quadtree and multi-type tree (QTMT) partition structure with nested multi-class trees so that the coding unit (CU) partition can better match the video texture features. This partition structure makes the compression efficiency of VVC significantly improved, but the computational complexity is also significantly increased, resulting in an increase in encoding time. Therefore, we propose a fast CU partition decision algorithm based on DenseNet network and decision tree (DT) classifier to reduce the coding complexity of VVC and save more coding time. We extract spatial feature vectors based on the DenseNet network model. Spatial feature vectors are constructed by predicting the boundary probabilities of 4 × 4 blocks in 64 × 64 coding units. Then, using the spatial features as the input of the DT classifier, through the classification function of the DT classifier model, the top N division modes with higher prediction probability are selected, and other division modes are skipped to reduce the computational complexity. Finally, the optimal partition mode is selected by comparing the RD cost. Our proposed algorithm achieves 47.6% encoding time savings on VTM10.0, while BDBR only increases by 0.91%.

show abstract

Fast Algorithm for HEVC using Frequency-based Convolutional Neural Networks

Wu,

Chen

2023

2023 3rd International Conference on Electronic Information Engineering and Computer (EIECT)

View full text Add to dashboard Cite

A CNN-based QTMT partitioning decision for the VVC standard

Cited by 3 publications

References 9 publications

A Fast Decision Algorithm for VVC Intra-Coding Based on Texture Feature and Machine Learning

A Fast Decision Algorithm for VVC Intra-Coding Based on Texture Feature and Machine Learning

Fast CU Decision Algorithm Based on CNN and Decision Trees for VVC

Fast Algorithm for HEVC using Frequency-based Convolutional Neural Networks

Contact Info

Product

Resources

About