Ismael Seidel
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Ismael Seidel has a total of 34 co-authors. Here are the top co-authors by number of shared publications:
- José Luís Güntzel
UFSC - Universidade Federal de Santa Catarina
5 shared publications - Luiz Henrique Cancellier
Department of Informatics and Statistics (INE/PPGCC), Federal University of Santa Catarina (UFSC), Florianpolis, Brazil
3 shared publications - André Beims Bräscher
Department of Informatics and Statistics (INE/PPGCC), Federal University of Santa Catarina (UFSC), Florianpolis, Brazil
2 shared publications Jose Luis Guntzel
Embedded Computing Lab. (ECL), Dept. of Computer Science and Statistics (INE-PPGCC), Federal University of Santa Catarina (UFSC), Florianópolis, Brazil
2 shared publications- Luciano Agostini
UFPEL
2 shared publications Marcio Monteiro
Embedded Computing Lab. (ECL), Dept. of Computer Science and Statistics (INE-PPGCC), Federal University of Santa Catarina (UFSC), Florianópolis, Brazil
2 shared publications
Latest Publications
- Open Access
- 30 December 2025
- Published by Journal of Integrated Circuits and Systems in Journal of Integrated Circuits and Systems
- Vol. 20 (3) , 1-13
DIVC-Based Pseudo-Temporal Light Field Coding
The JPEG Pleno Light Field has been developed to fill the gap in standardized methods for Light Field (LF) coding.Alternatively, LFs can be represented as pseudo-temporal sequences, unlocking the myriad of video coding techniques and standards for LF coding.Recently, the Decoupled Interpolated Video Coding (DIVC) method has been proposed, combining novel Neural Network-based techniques with hybrid video coding techniques.Such method eliminates frames from a video sequence being encoded and then, on the decoder side, it employs Video Frame Interpolation (VFI) as a way of regenerating removed frames.In this work, we demonstrate the application of the DIVC for Light Field Coding (DIVC-LF) method as a way of efficiently coding LF data while also relying on well established and optimized hybrid video coding techniques.According to our experiments, DIVC-LF improves the coding efficiency compared to the state-of-the-art VVC Test Model (VTM).On average, our method achieves BD-BR of -5.44% up to -28.93%, depending on different combinations of coding configuration, LF dataset, and quality metrics.The DIVC-LF also provides better coding efficiency compared to JPEG Pleno 4DTM by achieving, on average, BD-BR between -45.92% and -59.74%, depending on coding configuration.
- 30 December 2025
- Published by Journal of Integrated Circuits and Systems in Journal of Integrated Circuits and Systems
- Vol. 20 (3) , 1-13
Open AccessDIVC-Based Pseudo-Temporal Light Field Coding
The JPEG Pleno Light Field has been developed to fill the gap in standardized methods for Light Field (LF) coding.Alternatively, LFs can be represented as pseudo-temporal sequences, unlocking the myriad of video coding techniques and standards for LF coding.Recently, the Decoupled Interpolated Video Coding (DIVC) method has been proposed, combining novel Neural Network-based techniques with hybrid video coding techniques.Such method eliminates frames from a video sequence being encoded and then, on the decoder side, it employs Video Frame Interpolation (VFI) as a way of regenerating removed frames.In this work, we demonstrate the application of the DIVC for Light Field Coding (DIVC-LF) method as a way of efficiently coding LF data while also relying on well established and optimized hybrid video coding techniques.According to our experiments, DIVC-LF improves the coding efficiency compared to the state-of-the-art VVC Test Model (VTM).On average, our method achieves BD-BR of -5.44% up to -28.93%, depending on different combinations of coding configuration, LF dataset, and quality metrics.The DIVC-LF also provides better coding efficiency compared to JPEG Pleno 4DTM by achieving, on average, BD-BR between -45.92% and -59.74%, depending on coding configuration.
- Open Access
- 24 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 73 (3) , 1565-1578
A Parallel JPEG Pleno Baseline Block-Based Profile Light Field Codec Using OpenMP
Light Fields (LFs) are an image modality derived from the plenoptic function, capable of representing scenes with a richer amount of visual information, making them well-suitable for immersive applications. To efficiently compress this data-intensive modality, the Joint Photographic Experts Group (JPEG) committee created the JPEG Pleno Part 2 standard with two profiles. This work focuses on the reference encoder and decoder implementations for the Baseline Block-Based Profile (BBBP), i.e., the JPEG PLeno Model (JPLM). Our main contribution lies in the proposal and analysis of a parallel implementation of JPLM codec using OpenMP, including the coding efficiency overhead of three solutions to support parallel decoding: two based on a marker defined in Part 2, called PNT; and another non-compliant version of the entropy codec. We show that it is possible to accelerate encoding up to $12\times $ when using 24threads, and decoding up to $4\times $ with 10 threads. In both cases, the memory overhead remained below 20% for the tested LFs. Although the parallel encoder produces bit-exact matches when compared to the sequential version, a PNT marker is required to ensure parallel decoding, which causes an increase of about 4% in terms of BD-Rate. On the other hand, we show that by updating the arithmetic codec to avoid the PNT, this overhead reduces to at most 0.22%.
- 24 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 73 (3) , 1565-1578
Open AccessA Parallel JPEG Pleno Baseline Block-Based Profile Light Field Codec Using OpenMP
Light Fields (LFs) are an image modality derived from the plenoptic function, capable of representing scenes with a richer amount of visual information, making them well-suitable for immersive applications. To efficiently compress this data-intensive modality, the Joint Photographic Experts Group (JPEG) committee created the JPEG Pleno Part 2 standard with two profiles. This work focuses on the reference encoder and decoder implementations for the Baseline Block-Based Profile (BBBP), i.e., the JPEG PLeno Model (JPLM). Our main contribution lies in the proposal and analysis of a parallel implementation of JPLM codec using OpenMP, including the coding efficiency overhead of three solutions to support parallel decoding: two based on a marker defined in Part 2, called PNT; and another non-compliant version of the entropy codec. We show that it is possible to accelerate encoding up to $12\times $ when using 24threads, and decoding up to $4\times $ with 10 threads. In both cases, the memory overhead remained below 20% for the tested LFs. Although the parallel encoder produces bit-exact matches when compared to the sequential version, a PNT marker is required to ensure parallel decoding, which causes an increase of about 4% in terms of BD-Rate. On the other hand, we show that by updating the arithmetic codec to avoid the PNT, this overhead reduces to at most 0.22%.
- Open Access
- 17 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 31671-31682
Asymmetric Multi-Layer Compression: Decoupling Inter-Layer Coding Dependencies With a Learned Model
Multi-layer compression is employed to achieve scalability by providing a base bitstream and additional enhancement layers to be used when extra bandwidth is available. In this work, we combine hand-crafted and learned solutions, isolating the former in the base layer and using the latter in the enhancement layer. Moreover, the proposed Asymmetric Multi-layer Compression (AMLC) model decouples the base and enhancement layers at the encoding side, leveraging the end-to-end enhancement layer model to simplify the coding process. AMLC encodes an image 1.28× faster than a fully learned multi-layer codec while maintaining similar quality and coding efficiency despite the lower quality of the handcrafted base layer used in AMLC. While the proposed asymmetric encoder together with the handcrafted base layer reduce the overall complexity, the learned enhancement layer brings benefits in terms of coding efficiency. AMLC outperforms the coding efficiency of the Scalable HEVC (SHVC) reference software, a handcrafted multi-layer codec, achieving BD-Rate gains of -27.86%, -34.19%, and -26.57% for PSNR-YUV, PSNR-RGB, and MS-SSIM quality metrics, respectively. Finally, an equivalent implementation of the proposed model adding the inter-layer dependency on the encoding side produces results that are similar to the ones from AMLC’s, confirming its potential to simplify multi-layer compression without significant losses in coding efficiency.
- 17 February 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 13, 31671-31682
Open AccessAsymmetric Multi-Layer Compression: Decoupling Inter-Layer Coding Dependencies With a Learned Model
Multi-layer compression is employed to achieve scalability by providing a base bitstream and additional enhancement layers to be used when extra bandwidth is available. In this work, we combine hand-crafted and learned solutions, isolating the former in the base layer and using the latter in the enhancement layer. Moreover, the proposed Asymmetric Multi-layer Compression (AMLC) model decouples the base and enhancement layers at the encoding side, leveraging the end-to-end enhancement layer model to simplify the coding process. AMLC encodes an image 1.28× faster than a fully learned multi-layer codec while maintaining similar quality and coding efficiency despite the lower quality of the handcrafted base layer used in AMLC. While the proposed asymmetric encoder together with the handcrafted base layer reduce the overall complexity, the learned enhancement layer brings benefits in terms of coding efficiency. AMLC outperforms the coding efficiency of the Scalable HEVC (SHVC) reference software, a handcrafted multi-layer codec, achieving BD-Rate gains of -27.86%, -34.19%, and -26.57% for PSNR-YUV, PSNR-RGB, and MS-SSIM quality metrics, respectively. Finally, an equivalent implementation of the proposed model adding the inter-layer dependency on the encoding side produces results that are similar to the ones from AMLC’s, confirming its potential to simplify multi-layer compression without significant losses in coding efficiency.
- 28 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
A New Approach to Video Coding Leveraging Hybrid Coding and Video Frame Interpolation
In this work, we propose the use of a video coding method, dubbed Decoupled Interpolated Video Coding (DIVC), which blends traditional hybrid video coding with novel approaches based on Neural Networks (NNs). The DIVC approach provides a base-level representation with a reduced bit rate by dropping frames in a regular manner which can be decoded by standard hybrid video coding. Meanwhile, we also regenerate the dropped frames from the reconstructed video from the base-level representation using NN-based Video Frame Interpolation (VFI) to recover the original number of frames per second (fps). We show that the DIVC approach can improve video coding efficiency considering the perceptual quality measured with the Structural Similarity (SSIM) metric. More specifically, such approach achieved a reduction of 38.96% (AI configuration) and 5% (RA configuration) in terms of SSIM dB -based Bøntegaard Delta Bit Rate (BD-Rate) when compared to traditional video coding. Meanwhile, considering the Peak Signal-to-Noise Ratio (PSNR), which may not correlate as well to human perception, the discussed approach resulted in PSNR BD-Rate reduction of 16.19% in AI configuration and increase of 19.33% in RA configuration. Employing VFI to improve video coding efficiency tends to be more suited to processing video sequences with high fps or slower motion.
- 28 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
A New Approach to Video Coding Leveraging Hybrid Coding and Video Frame Interpolation
In this work, we propose the use of a video coding method, dubbed Decoupled Interpolated Video Coding (DIVC), which blends traditional hybrid video coding with novel approaches based on Neural Networks (NNs). The DIVC approach provides a base-level representation with a reduced bit rate by dropping frames in a regular manner which can be decoded by standard hybrid video coding. Meanwhile, we also regenerate the dropped frames from the reconstructed video from the base-level representation using NN-based Video Frame Interpolation (VFI) to recover the original number of frames per second (fps). We show that the DIVC approach can improve video coding efficiency considering the perceptual quality measured with the Structural Similarity (SSIM) metric. More specifically, such approach achieved a reduction of 38.96% (AI configuration) and 5% (RA configuration) in terms of SSIM dB -based Bøntegaard Delta Bit Rate (BD-Rate) when compared to traditional video coding. Meanwhile, considering the Peak Signal-to-Noise Ratio (PSNR), which may not correlate as well to human perception, the discussed approach resulted in PSNR BD-Rate reduction of 16.19% in AI configuration and increase of 19.33% in RA configuration. Employing VFI to improve video coding efficiency tends to be more suited to processing video sequences with high fps or slower motion.
- 1 February 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 31 (10577149) , 1994-2003
Differential Transform for Video-Based Plenoptic Point Cloud Coding
Point cloud compression has been studied in standard bodies and we are here concerned with the Moving Picture Experts Group video-based point cloud compression (V-PCC) solution. Plenoptic point clouds (PPC) is a novel volumetric data representation wherein points are associated with colors in all viewing directions to improve realism. It is sampled as a number ( $N_{c}$ ) of attribute colors per point. We propose a new method for the efficient video-based compression of PPC that is backwards compatible with the existing single-color V-PCC decoder. V-PCC generates three image atlases which are encoded using an image/video encoder. We assume there may be a reference color which is to be encoded as the main payload. We generate $N_{c}+3$ atlases and we produce $N_{c}$ differential images against the reference color image. Those difference images are pixel-wise transformed using an $N_{c}$ -point discrete cosine transform, generating $N_{c}$ transformed atlases which are encoded, forming the secondary payload. Such secondary information is the plenoptic enhancement to the point cloud. If there is no reference attribute, we skip the differences and use the lowest frequency of the transformed atlases as the main payload. Results are presented that show an unrivaled performance of the proposed method.
- 1 February 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Image Processing
- Vol. 31 (10577149) , 1994-2003
Differential Transform for Video-Based Plenoptic Point Cloud Coding
Point cloud compression has been studied in standard bodies and we are here concerned with the Moving Picture Experts Group video-based point cloud compression (V-PCC) solution. Plenoptic point clouds (PPC) is a novel volumetric data representation wherein points are associated with colors in all viewing directions to improve realism. It is sampled as a number ( $N_{c}$ ) of attribute colors per point. We propose a new method for the efficient video-based compression of PPC that is backwards compatible with the existing single-color V-PCC decoder. V-PCC generates three image atlases which are encoded using an image/video encoder. We assume there may be a reference color which is to be encoded as the main payload. We generate $N_{c}+3$ atlases and we produce $N_{c}$ differential images against the reference color image. Those difference images are pixel-wise transformed using an $N_{c}$ -point discrete cosine transform, generating $N_{c}$ transformed atlases which are encoded, forming the secondary payload. Such secondary information is the plenoptic enhancement to the point cloud. If there is no reference attribute, we skip the differences and use the lowest frequency of the transformed atlases as the main payload. Results are presented that show an unrivaled performance of the proposed method.
- Open Access
- 18 September 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 170807-170829
The JPEG Pleno Light Field Coding Standard 4D-Transform Mode: How to Design an Efficient 4D-Native Codec
The increasing demand for highly realistic and immersive visual experiences has led to the emergence of richer 3D visual representation models such as light fields, point clouds and meshes. Light fields may be modelled as a 2D array of 2D views, corresponding to a large amount of data, which demands for highly efficient coding solutions. Although static light fields are inherently 4D structures, the coding solutions in the literature mostly employ traditional 2D coding tools associated with techniques such as depth-based image rendering to generate a residual, again to be coded using available 2D coding tools. To address the market needs, JPEG has launched the so-called JPEG Pleno standard, which Part 2 is dedicated to light field coding. The JPEG Pleno Light Field Coding standard includes two coding modes, one based on the 4D-DCT, so-called 4D-Transform mode, and another based on depth-based synthesis, so-called 4D-Prediction. The 4D-Transform coding mode standardizes for the first time a 4D-native light field coding solution where the full light field redundancy across the four dimensions is comprehensively exploited, somehow extending to 4D the 2D coding framework adopted decades ago by the popular JPEG Baseline standard. In this context, this paper describes and analyzes in detail the conceptual and algorithmic design process which has led to the creation of the JPEG Pleno Light Field Coding standard 4D-Transform coding mode. This has happened through a sequence of steps involving technical innovation design and integration where increasingly sophisticated coding tools have been combined and improved to maximize the final rate distortion (RD) performance.
- 18 September 2020
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 8, 170807-170829
Open AccessThe JPEG Pleno Light Field Coding Standard 4D-Transform Mode: How to Design an Efficient 4D-Native Codec
The increasing demand for highly realistic and immersive visual experiences has led to the emergence of richer 3D visual representation models such as light fields, point clouds and meshes. Light fields may be modelled as a 2D array of 2D views, corresponding to a large amount of data, which demands for highly efficient coding solutions. Although static light fields are inherently 4D structures, the coding solutions in the literature mostly employ traditional 2D coding tools associated with techniques such as depth-based image rendering to generate a residual, again to be coded using available 2D coding tools. To address the market needs, JPEG has launched the so-called JPEG Pleno standard, which Part 2 is dedicated to light field coding. The JPEG Pleno Light Field Coding standard includes two coding modes, one based on the 4D-DCT, so-called 4D-Transform mode, and another based on depth-based synthesis, so-called 4D-Prediction. The 4D-Transform coding mode standardizes for the first time a 4D-native light field coding solution where the full light field redundancy across the four dimensions is comprehensively exploited, somehow extending to 4D the 2D coding framework adopted decades ago by the popular JPEG Baseline standard. In this context, this paper describes and analyzes in detail the conceptual and algorithmic design process which has led to the creation of the JPEG Pleno Light Field Coding standard 4D-Transform coding mode. This has happened through a sequence of steps involving technical innovation design and integration where increasingly sophisticated coding tools have been combined and improved to maximize the final rate distortion (RD) performance.
- 19 March 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 66 (6) , 2102-2115
Energy-Efficient Hadamard-Based SATD Hardware Architectures Through Calculation Reuse
The Hadamard-based Sum of Absolute Transformed Differences (SATD) is a distortion metric that correlates better with other video encoding steps than the commonly used Sum of Absolute Differences (SAD) and thus potentially improves coding efficiency. However, due to the Hadamard Transform (HT), it demands a large share of total encoding time. Meanwhile, the widespread use of embedded encoders made computational complexity and energy efficiency two relevant issues, besides the usual trade-off between the rate and distortion. Hence, the adoption of SATD during video coding requires the design of energy-efficient hardware accelerators. In this paper, we show that the SATD improves coding efficiency by up to -2.99% in BD-Rate when used in motion estimation within the HEVC Model (HM). Subsequently, as the main contribution, we combined two properties featured by the SATD to devise a novel hardware architecture based on calculation reuse. These properties allow for rearranging the architecture components to reduce the hardware size, save operations, and thus, reduce energy. The synthesis and simulation results indicate that the proposed architecture can lead to area and energy savings of about 28% and 32%, respectively, compared to the state-of-the-art architecture based on reuse solely.
- 19 March 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 66 (6) , 2102-2115
Energy-Efficient Hadamard-Based SATD Hardware Architectures Through Calculation Reuse
The Hadamard-based Sum of Absolute Transformed Differences (SATD) is a distortion metric that correlates better with other video encoding steps than the commonly used Sum of Absolute Differences (SAD) and thus potentially improves coding efficiency. However, due to the Hadamard Transform (HT), it demands a large share of total encoding time. Meanwhile, the widespread use of embedded encoders made computational complexity and energy efficiency two relevant issues, besides the usual trade-off between the rate and distortion. Hence, the adoption of SATD during video coding requires the design of energy-efficient hardware accelerators. In this paper, we show that the SATD improves coding efficiency by up to -2.99% in BD-Rate when used in motion estimation within the HEVC Model (HM). Subsequently, as the main contribution, we combined two properties featured by the SATD to devise a novel hardware architecture based on calculation reuse. These properties allow for rearranging the architecture components to reduce the hardware size, save operations, and thus, reduce energy. The synthesis and simulation results indicate that the proposed architecture can lead to area and energy savings of about 28% and 32%, respectively, compared to the state-of-the-art architecture based on reuse solely.
- 1 February 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
On the calculation reuse in hadamard-based SATD
The Sum of Absolute Transformed Differences (SATD) is a distortion metric able to improve video coding efficiency because of its correlation with the transform step of hybrid video encoders. For the sake of energy efficiency, and also because it demands a large share of the total encoding time, SATD is a natural candidate for hardware implementation. One way to reduce power is through calculation reuse. In this paper, we prove that previously computed Hadamard Transform (HT) can be re-used in the computation of larger blocks. Also, we propose a generic SATD hardware design which implements such reuse. We evaluated our proposed design in terms of area, power and energy. Compared to state-of-the-art SATD designs, our solution is about 45% more energy-efficient.
- 1 February 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-4
On the calculation reuse in hadamard-based SATD
The Sum of Absolute Transformed Differences (SATD) is a distortion metric able to improve video coding efficiency because of its correlation with the transform step of hybrid video encoders. For the sake of energy efficiency, and also because it demands a large share of the total encoding time, SATD is a natural candidate for hardware implementation. One way to reduce power is through calculation reuse. In this paper, we prove that previously computed Hadamard Transform (HT) can be re-used in the computation of larger blocks. Also, we propose a generic SATD hardware design which implements such reuse. We evaluated our proposed design in terms of area, power and energy. Compared to state-of-the-art SATD designs, our solution is about 45% more energy-efficient.
- 1 February 2016
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 327-330
Squarer exploration for energy-efficient sum of squared differences
The main reason for the long time and high energy requirements of state-of-the-art Video Coding (VC) standards, such as the HEVC, is the large amount of distortion calculations. Among the most known and used ones is the Sum of Squared Differences (SSD) which has a strong correlation with the Peak Signal-to-Noise Ratio (PSNR). Such correlation is explored by current encoders to provide a good trade-off between rate and distortion. Once VC is mandatory in current battery-powered devices, the adopted distortion metric must be as energy-efficient as possible. Although simple, the SSD requires a square operation, which hardware realization is costly. Thus, some VC hardware designs replace the SSD by the Sum of Absolute Differences (SAD). However, using SAD instead of SSD pays a price in coding efficiency. In this work we investigate four hardware designs for the square operation. Synthesis results for the designed architectures are compared to a reference SAD design from the literature. The best SSD architecture, using clock gating, requires only 20% more energy than SAD.
- 1 February 2016
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 327-330
Squarer exploration for energy-efficient sum of squared differences
The main reason for the long time and high energy requirements of state-of-the-art Video Coding (VC) standards, such as the HEVC, is the large amount of distortion calculations. Among the most known and used ones is the Sum of Squared Differences (SSD) which has a strong correlation with the Peak Signal-to-Noise Ratio (PSNR). Such correlation is explored by current encoders to provide a good trade-off between rate and distortion. Once VC is mandatory in current battery-powered devices, the adopted distortion metric must be as energy-efficient as possible. Although simple, the SSD requires a square operation, which hardware realization is costly. Thus, some VC hardware designs replace the SSD by the Sum of Absolute Differences (SAD). However, using SAD instead of SSD pays a price in coding efficiency. In this work we investigate four hardware designs for the square operation. Synthesis results for the designed architectures are compared to a reference SAD design from the literature. The best SSD architecture, using clock gating, requires only 20% more energy than SAD.
- 20 June 2015
- Published by Springer Nature in Analog Integrated Circuits and Signal Processing
- Vol. 85 (1) , 107-128
Towards optimal use of pel decimation to trade off quality for energy
- 20 June 2015
- Published by Springer Nature in Analog Integrated Circuits and Signal Processing
- Vol. 85 (1) , 107-128
Towards optimal use of pel decimation to trade off quality for energy