Cuiling Lan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Cuiling Lan has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
Feng Wu
Microsoft Research Asia, Beijing, China
2 shared publicationsGuangming Shi
School of Electronic Engineering, Xidian University, Xi'an, China
2 shared publicationsJizheng Xu
Microsoft Research Asia, Beijing, China
1 shared publication
Latest Publications
- 18 August 2011
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Signal Processing
- Vol. 5 (7) , 1298-1308
Exploiting Non-Local Correlation via Signal-Dependent Transform (SDT)
Over the past few decades, many studies on image and video compression have found various approaches to the exploitation of spatial and temporal local correlations. However, we believe it is imperative to find more efficient methods to progress the development of image and video compression. In this paper, we first study spatial non-local correlation, deducing that there exist strong correlations in non-local regions. However, it is rather difficult to make use of these non-local correlations while simultaneously minimizing overhead. To solve this problem, we propose the signal-dependent transform (SDT), which is derived from decoded non-local blocks that are selected by matching neighboring pixels. Since the encoder and decoder can use the same methods to derive the proposed transform, we can successfully eliminate overhead. Finally, we have implemented the proposed transform into the Key Technology Area (KTA) software to exploit both spatial and temporal non-local correlations. The experimental results show that the coding gain over KTA can be as high as 1.4 dB in intra-frame coding, and up to 1.0 dB in inter-frame coding. We believe we have effectively created an alternate method to improve image and video compression.
- 18 August 2011
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Selected Topics in Signal Processing
- Vol. 5 (7) , 1298-1308
Exploiting Non-Local Correlation via Signal-Dependent Transform (SDT)
Over the past few decades, many studies on image and video compression have found various approaches to the exploitation of spatial and temporal local correlations. However, we believe it is imperative to find more efficient methods to progress the development of image and video compression. In this paper, we first study spatial non-local correlation, deducing that there exist strong correlations in non-local regions. However, it is rather difficult to make use of these non-local correlations while simultaneously minimizing overhead. To solve this problem, we propose the signal-dependent transform (SDT), which is derived from decoded non-local blocks that are selected by matching neighboring pixels. Since the encoder and decoder can use the same methods to derive the proposed transform, we can successfully eliminate overhead. Finally, we have implemented the proposed transform into the Key Technology Area (KTA) software to exploit both spatial and temporal non-local correlations. The experimental results show that the coding gain over KTA can be as high as 1.4 dB in intra-frame coding, and up to 1.0 dB in inter-frame coding. We believe we have effectively created an alternate method to improve image and video compression.
- 15 December 2009
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 19 (4) , 946-957
Compress Compound Images in H.264/MPGE-4 AVC by Exploiting Spatial Correlation
Compound images are a combination of text, graphics and natural image. They present strong anisotropic features, especially on the text and graphics parts. These anisotropic features often render conventional compression inefficient. Thus, this paper proposes a novel coding scheme from the H.264 intraframe coding. In the scheme, two new intramodes are developed to better exploit spatial correlation in compound images. The first is the residual scalar quantization (RSQ) mode, where intrapredicted residues are directly quantized and coded without transform. The second is the base colors and index map (BCIM) mode that can be viewed as an adaptive color quantization. In this mode, an image block is represented by several representative colors, referred to as base colors, and an index map to compress. Every block selects its coding mode from two new modes and the previous intramodes in H.264 by rate-distortion optimization (RDO). Experimental results show that the proposed scheme improves the coding efficiency even more than 10 dB at most bit rates for compound images and keeps a comparable efficient performance to H.264 for natural images.
- 15 December 2009
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 19 (4) , 946-957
Compress Compound Images in H.264/MPGE-4 AVC by Exploiting Spatial Correlation
Compound images are a combination of text, graphics and natural image. They present strong anisotropic features, especially on the text and graphics parts. These anisotropic features often render conventional compression inefficient. Thus, this paper proposes a novel coding scheme from the H.264 intraframe coding. In the scheme, two new intramodes are developed to better exploit spatial correlation in compound images. The first is the residual scalar quantization (RSQ) mode, where intrapredicted residues are directly quantized and coded without transform. The second is the base colors and index map (BCIM) mode that can be viewed as an adaptive color quantization. In this mode, an image block is represented by several representative colors, referred to as base colors, and an index map to compress. Every block selects its coding mode from two new modes and the previous intramodes in H.264 by rate-distortion optimization (RDO). Experimental results show that the proposed scheme improves the coding efficiency even more than 10 dB at most bit rates for compound images and keeps a comparable efficient performance to H.264 for natural images.