A Novel Framework for Designing Directional Linear Transforms with Application to Video Compression

Masera, Maurizio; Fracastoro, Giulia; Martina, Maurizio; Magli, Enrico

doi:10.1109/icassp.2019.8683527

Cited by 4 publications

(15 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, removing the hardware support for the steering part with N = 32 (SDCT-16), the area becomes comparable with the one of the DCT. As a consequence, this solution can be of interest to increase the rate-distortion performance [8].…”

Section: Resultsmentioning

confidence: 99%

“…The complexity of a transform that is not separable is far greater than a separable one, so this may be a big drawback for the implementation. However, the complexity can be decreased drastically by splitting the SDCT into two parts, namely, a separable 2D DCT followed by some rotations, and then by computing the separable transform before applying rotations, as reported in [8]:…”

Section: Datapathmentioning

confidence: 99%

“…in Figure 6 are implemented as shift and add, as the number of possible rotation angles has been fixed to 8 (from 0, no rotation, to 7), as reported as optimum in [8]…”

Section: Datapathmentioning

confidence: 99%

“…This motivated several researchers to propose dedicated architectures for variable size DCTs, such as [4][5][6]. Recently, G. Fracastoro et al [7] proposed the Steerable DCT (SDCT) and showed that it can give some coding advantages when embedded in the HEVC standard [8]. Such a directional transform is not tailored to any specific one, but it can be potentially applied to any two-dimensional separable transform.…”

Section: Introductionmentioning

confidence: 99%

“…The SDCT kernels still retain a square shape so that computation remains easy to perform, even though this 2-D transform is not separable in two 1-D operations as for the classic 2-D DCT. Lately, the work in [8] demonstrated that it is possible to split the steerable cosine transform into a traditional DCT followed by a geometrical rotation. The resulting kernels are the same as the SDCT but the computation workload is reduced by exploiting the 2-D DCT separability.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Steerable-Discrete-Cosine-Transform (SDCT): Hardware Implementation and Performance Analysis

Peloso

Capra

Sole

et al. 2020

Sensors

Self Cite

View full text Add to dashboard Cite

In the last years, the need for new efficient video compression methods grown rapidly as frame resolution has increased dramatically. The Joint Collaborative Team on Video Coding (JCT-VC) effort produced in 2013 the H.265/High Efficiency Video Coding (HEVC) standard, which represents the state of the art in video coding standards. Nevertheless, in the last years, new algorithms and techniques to improve coding efficiency have been proposed. One promising approach relies on embedding direction capabilities into the transform stage. Recently, the Steerable Discrete Cosine Transform (SDCT) has been proposed to exploit directional DCT using a basis having different orientation angles. The SDCT leads to a sparser representation, which translates to improved coding efficiency. Preliminary results show that the SDCT can be embedded into the HEVC standard, providing better compression ratios. This paper presents a hardware architecture for the SDCT, which is able to work at a frequency of 188 M Hz , reaching a throughput of 3.00 GSample/s. In particular, this architecture supports 8k UltraHigh Definition (UHD) (7680 × 4320) with a frame rate of 60 Hz , which is one of the best resolutions supported by HEVC.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Datapathmentioning

confidence: 99%

“…in Figure 6 are implemented as shift and add, as the number of possible rotation angles has been fixed to 8 (from 0, no rotation, to 7), as reported as optimum in [8]…”

Section: Datapathmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Steerable-Discrete-Cosine-Transform (SDCT): Hardware Implementation and Performance Analysis

Peloso

Capra

Sole

et al. 2020

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

VLSI Architectures for the Steerable-Discrete-Cosine-Transform (SDCT)

Sole

Peloso

Capra

et al. 2020

Lecture Notes in Electrical Engineering

View full text Add to dashboard Cite

Image Compression Based on a Partially Rotated Discrete Cosine Transform With a Principal Orientation

Lee

Choe

2021

IEEE Access

View full text Add to dashboard Cite

Image transforms are necessary for image and video compression. Analytic transforms are powerful in compacting natural signals for wider exploitation. Various methods have been introduced to represent such data as a small number of bases, and several of these methods use machine learning, usually based on sparse coding, to outperform analytic transforms. They show sufficient data compaction abilities. However, these methods focus only on data compaction and reconstruction performance, without considering computational issues during implementation. We introduce a new framework for a more efficient transform based on a two-dimensional discrete cosine transform (DCT) and its characteristics. We aimed to improve the data compaction ability of transforms to levels better or similar to that of the DCT and other data-driven transforms, with fast and efficient implementation. We focused on the properties of the DCT, including horizontal and vertical directional information, and approximated its direction using the transform. Our framework was designed by rotating some of the DCT bases to fit this direction. As expected, our framework achieves a transform design with minimized computation for efficient implementation. It does not require an iterative algorithm or brute-force methods to find the best transform matrix or other parameters, thereby making it much faster than other methods. Our framework is 10 times faster than the steerable DCT (SDCT) and twice as fast as the eight-level SDCT with minimum performance reduction. Experimental validation with various images indicates that the proposed method sufficiently approaches the performance of the other transforms despite faster implementation.INDEX TERMS Directional discrete cosine transform, discrete cosine transform, image transformation, image compression, sparse coding transform.

show abstract

A Novel Framework for Designing Directional Linear Transforms with Application to Video Compression

Cited by 4 publications

References 17 publications

Steerable-Discrete-Cosine-Transform (SDCT): Hardware Implementation and Performance Analysis

Steerable-Discrete-Cosine-Transform (SDCT): Hardware Implementation and Performance Analysis

VLSI Architectures for the Steerable-Discrete-Cosine-Transform (SDCT)

Image Compression Based on a Partially Rotated Discrete Cosine Transform With a Principal Orientation

Contact Info

Product

Resources

About