Image & Video Processing

Marcos V. Conde, Gregor Geigle, Radu Timofte

This work addresses the challenge of flexible and high-quality image restoration for users by enabling natural language control, representing a novel benchmark in the field.

Xinjie Zhang, Xingtong Ge, Tongda Xu et al.

This addresses the limitation of INRs on low-end devices by offering a more efficient alternative for image processing applications.

Ibrahim Ethem Hamamci, Sezgin Er, Bjoern Menze

This addresses the need to reduce radiologists' workload by extending automated report generation to 3D imaging, which is an incremental advancement from existing 2D methods.

Zefan Yang, Xuanang Xu, Jiajin Zhang et al.

This work addresses the problem of developing a generalizable and efficient chest X-ray classification model for the medical community, particularly for tasks with limited labeled data.

Reuben Dorent, Erickson Torio, Nazim Haouchine et al.

This addresses the problem of improving surgical outcomes for neurosurgeons by enabling more accurate and adaptable tumor segmentation during brain surgery.

Vanshali Sharma, Debesh Jha, M. K. Bhuyan et al.

This work addresses the problem of accurate polyp classification for clinicians and medical researchers, particularly in the context of colorectal cancer diagnosis, by providing a more robust and efficient image analysis tool.

Chicago Y. Park, Yuyang Hu, Michael T. McCann et al.

This provides a novel perspective for researchers in computational imaging, allowing direct comparison between PnP and SBM-based reconstruction methods using the same neural network prior.

Georgy Perevozchikov, Nancy Mehta, Mahmoud Afifi et al.

This addresses the problem of adapting learnable ISPs to new camera models without paired data for smartphone photography.

Tongda Xu, Ziran Zhu, Dailan He et al.

This addresses perceptual image compression for applications requiring high-quality reconstructions, offering a novel theoretical equivalence that simplifies implementation.

Hao Shi, Song Wang, Jiaming Zhang et al.

This work addresses the problem of accurate 3D scene understanding for autonomous vehicles, offering a novel offboard solution that improves upon existing vision-based methods.

Mang Ning, Mingxiao Li, Jianlin Su et al.

This addresses image generation efficiency for AI applications, offering a novel approach with significant computational savings.

Yulin Wang, Yang Yue, Yang Yue et al. · tsinghua

This addresses the resource-intensive limitations of computer vision for applications like autonomous driving and medical imaging, offering a paradigm shift toward more efficient and flexible models.

Syed Ariff Syed Hesham, Yun Liu, Guolei Sun et al.

This work addresses the challenge of limited temporal context and high computational costs in video semantic segmentation, offering a more efficient solution for applications like autonomous driving or video analysis.

Yuzhi Zhao, Lai-Man Po, Xin Ye et al.

This addresses the persistent challenge of image degradation from noise and blur in imaging systems, offering a solution that avoids trade-offs in single-image methods and misalignment issues in multi-image approaches.

Xuanyu Tian, Lixuan Chen, Qing Wu et al.

This work addresses the problem of sparse-view computed tomography reconstruction for medical imaging applications, providing an unsupervised method that can handle noisy data and outperform supervised methods in certain scenarios.

Wen Yan, Qianye Yang, Shiqi Huang et al.

This provides a training-free, generalizable solution for image registration tasks, particularly beneficial for medical imaging where data curation is costly.

Matthieu Terris, Ulugbek S. Kamilov, Thomas Moreau

This work addresses a fundamental problem in imaging inverse problems for researchers and practitioners by offering a new paradigm for incorporating pretrained restoration models, though it is incremental as it builds on existing Plug-and-Play frameworks.

Vivek Gopalakrishnan, Neel Dey, David-Dimitris Chlorogiannis et al. · mit

This addresses the challenge of rapid and precise 2D/3D registration for surgical procedures, enabling broader clinical application without labor-intensive annotations.

Sidi Yang, Binxiao Huang, Mingdeng Cao et al.

This addresses the need for low-power, fast image enhancement on edge devices like smartphones and cameras, representing a novel approach rather than an incremental improvement.

Samuel Garske, Konrad Heidler, Bradley Evans et al.

It addresses the urgent need for effective, low-cost hazard monitoring for environmental management, offering a generalizable solution with improved detection and mapping capabilities.