ELIC: Efficient Learned Image Compression with Unevenly Grouped Space-Channel Contextual Adaptive Coding

He, Dailan; Yang, Ziming; Peng, Weikun; Ma, Rui; Qin, Hongwei; Wang, Yan

doi:10.1109/cvpr52688.2022.00563

Cited by 106 publications

(45 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These invalid pixels are caused by the absence of LiDAR laser signal return. In our prediction design, we leverage this property The residual coding network g a and g s consist of stacks of convolutions, residual blocks [9] and attention blocks [10]. For more details about the hyperprior model, please refer to [11].…”

Section: Inter-frame Predictionmentioning

confidence: 99%

“…The prior handcrafted methods are not capable of fully exploiting redundancy, leading to a poor compression ratio. Our learning-based approach is motivated by the recent research on learned color image compression [9,10,11,18]. A variational autoencoder (VAE) is adopted to achieve transform coding.…”

Section: Residual Frame Codingmentioning

confidence: 99%

“…(a) The prediction network takes decoded intra-frames xt−1 , xt+1 and the mask m t as inputs and generates a prediction xt . (b)The residual coding network g a and g s consist of stacks of convolutions, residual blocks[9] and attention blocks[10]. For more details about the hyperprior model, please refer to[11].…”

mentioning

confidence: 99%

See 2 more Smart Citations

University Technology Commercialization in Taiwan: National Taiwan University (NTU) and National Taiwan University of Science and Technology (NTUST)

Liu¹,

Chen²,

Chiou³

et al. 2011

Academic Entrepreneurship in Asia

View full text Add to dashboard Cite

The large amount of data collected by LiDAR sensors brings the issue of LiDAR point cloud compression (PCC). Previous works on LiDAR PCC have used range image representations and followed the predictive coding paradigm to create a basic prototype of a coding framework. However, their prediction methods give an inaccurate result due to the negligence of invalid pixels in range images and the omission of future frames in the time step. Moreover, their handcrafted design of residual coding methods could not fully exploit spatial redundancy. To remedy this, we propose a coding framework BIRD-PCC. Our prediction module is aware of the coordinates of invalid pixels in range images and takes a bidirectional scheme. Also, we introduce a deep-learned residual coding module that can further exploit spatial redundancy within a residual frame. Experiments conducted on SemanticKITTI and KITTI-360 datasets show that BIRD-PCC outperforms other methods in most bitrate conditions and generalizes well to unseen environments.

show abstract

Section: Inter-frame Predictionmentioning

confidence: 99%

Section: Residual Frame Codingmentioning

confidence: 99%

See 1 more Smart Citation

University Technology Commercialization in Taiwan: National Taiwan University (NTU) and National Taiwan University of Science and Technology (NTUST)

Liu¹,

Chen²,

Chiou³

et al. 2011

Academic Entrepreneurship in Asia

View full text Add to dashboard Cite

show abstract

“…Recently, end-to-end image compression has engaged increasing interests, which is built upon the transform coding paradigm with nonlinear transform and powerful entropy models for higher compression efficiency. Nonlinear transform is used to produce compact representations, such as generalized divisive normalization (GDN) (Ballé et al, 2015), the self-attention block (Cheng et al, 2020), wavelet-like invertible transform (Ma et al, 2020) and stacks of residual bottleneck blocks (He et al, 2022). To approximate the distribution of latent representations, many advanced entropy models have been proposed.…”

Section: Related Workmentioning

confidence: 99%

LDMIC: Learning-based Distributed Multi-view Image Coding

Zhang¹,

Shao²,

Zhang³

2023

Preprint

View full text Add to dashboard Cite

Multi-view image compression plays a critical role in 3D-related applications. Existing methods adopt a predictive coding architecture, which requires joint encoding to compress the corresponding disparity as well as residual information. This demands collaboration among cameras and enforces the epipolar geometric constraint between different views, which makes it challenging to deploy these methods in distributed camera systems with randomly overlapping fields of view. Meanwhile, distributed source coding theory indicates that efficient data compression of correlated sources can be achieved by independent encoding and joint decoding, which motivates us to design a learning-based distributed multi-view image coding (LDMIC) framework. With independent encoders, LDMIC introduces a simple yet effective joint context transfer module based on the cross-attention mechanism at the decoder to effectively capture the global inter-view correlations, which is insensitive to the geometric relationships between images. Experimental results show that LDMIC significantly outperforms both traditional and learningbased MIC methods while enjoying fast encoding speed. Code will be released at https://github.com/Xinjie-Q/LDMIC.

show abstract

“…In recent years, image coding methods based on end-toend optimization have been rapidly explored and developed, and show promise to become the next-generation coding standard. On the one hand, with the powerful image understanding and generation capabilities of deep learning, some very recent works outperform VVC in PSNR and MS-SSIM (Gao et al 2021;Guo et al 2021a;Xie, Cheng, and Chen 2021;Chen, Xu, and Wang 2022;He et al 2022). On the other hand, to meet the needs of industrial applications, researchers have designed flexible modules to implement variable bitrate (Cui et al 2021) and scalable coding (Guo, Zhang, and Chen 2019).…”

Section: Introductionmentioning

confidence: 99%

Improving Multi-generation Robustness of Learned Image Compression

Li¹,

Zheng²,

Wang³

2022

Preprint

View full text Add to dashboard Cite

Benefit from flexible network designs and end-to-end joint optimization approach, learned image compression (LIC) has demonstrated excellent coding performance and practical feasibility in recent years. However, existing compression models suffer from serious multi-generation loss, which always occurs during image editing and transcoding. During the process of repeatedly encoding and decoding, the quality of the image will rapidly degrade, resulting in various types of distortion, which significantly limits the practical application of LIC. In this paper, a thorough analysis is carried out to determine the source of generative loss in successive image compression (SIC). We point out and solve the quantization drift problem that affects SIC, reversibility loss function as well as channel relaxation method are proposed to further reduce the generation loss. Experiments show that by using our proposed solutions, LIC can achieve comparable performance to the first compression of BPG even after 50 times reencoding without any change of the network structure.

show abstract

ELIC: Efficient Learned Image Compression with Unevenly Grouped Space-Channel Contextual Adaptive Coding

Cited by 106 publications

References 24 publications

University Technology Commercialization in Taiwan: National Taiwan University (NTU) and National Taiwan University of Science and Technology (NTUST)

University Technology Commercialization in Taiwan: National Taiwan University (NTU) and National Taiwan University of Science and Technology (NTUST)

LDMIC: Learning-based Distributed Multi-view Image Coding

Improving Multi-generation Robustness of Learned Image Compression

Contact Info

Product

Resources

About