Rama Harihara
Research Subjects
Rama Harihara has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Sungye Kim
Advanced Micro Devices, Inc
3 shared publications - Wojciech Kaliński
Advanced Micro Devices, Inc
3 shared publications - Wojciech Uss
Advanced Micro Devices, Inc
3 shared publications - Alexandr Kuznetsov
Advanced Micro Devices, Inc
2 shared publications - Harish Anand
Advanced Micro Devices, Inc
2 shared publications - Damian Andrysiak
Advanced Micro Devices (AMD), Lodz, Poland
1 shared publication
Latest Publications
- 10 September 2026
- Published by Springer Nature
- p. 205-224
Temporally Stable Generative Illumination with a One-Step Diffusion Model
- 10 September 2026
- Published by Springer Nature
- p. 205-224
Temporally Stable Generative Illumination with a One-Step Diffusion Model
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-3
Lightweight Attention-based Indirect Illumination
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-3
Open AccessLightweight Attention-based Indirect Illumination
- Open Access
- 9 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-2
G-FED: G-Buffer Guided Frame Extrapolation in Video Diffusion Models
- 9 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 1-2
Open AccessG-FED: G-Buffer Guided Frame Extrapolation in Video Diffusion Models
- Open Access
- 22 May 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Computer Graphics and Interactive Techniques
- Vol. 8 (1) , 1-18
Joint Denoising and Upscaling via Multi-branch and Multi-scale Feature Network
Deep learning-based denoising and upscaling techniques have emerged to enhance framerates for real-time rendering. A single neural network for joint denoising and upscaling offers the advantage of sharing parameters in the feature space, enabling efficient prediction of filter weights for both. However, it is still ongoing research to devise an efficient feature extraction neural network that uses different characteristics in inputs for the two combined problems. We propose a multi-branch, multi-scale feature extraction network for joint neural denoising and upscaling. The proposed multi-branch U-Net architecture is lightweight and effectively accounts for different characteristics in noisy color and noise-free aliased auxiliary buffers. Our technique produces superior quality denoising in a target resolution (4K), given noisy 1spp Monte Carlo renderings and auxiliary buffers in a low resolution (1080p), compared to the state-of-the-art methods.
- 22 May 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Computer Graphics and Interactive Techniques
- Vol. 8 (1) , 1-18
Open AccessJoint Denoising and Upscaling via Multi-branch and Multi-scale Feature Network
Deep learning-based denoising and upscaling techniques have emerged to enhance framerates for real-time rendering. A single neural network for joint denoising and upscaling offers the advantage of sharing parameters in the feature space, enabling efficient prediction of filter weights for both. However, it is still ongoing research to devise an efficient feature extraction neural network that uses different characteristics in inputs for the two combined problems. We propose a multi-branch, multi-scale feature extraction network for joint neural denoising and upscaling. The proposed multi-branch U-Net architecture is lightweight and effectively accounts for different characteristics in noisy color and noise-free aliased auxiliary buffers. Our technique produces superior quality denoising in a target resolution (4K), given noisy 1spp Monte Carlo renderings and auxiliary buffers in a low resolution (1080p), compared to the state-of-the-art methods.