Speakers
Invited Speakers
We are honored to invite leading researchers from academia and industry to share their insights at the workshop.
Join us to explore the AGI-oriented future where generative models, reward learning, and autonomous agents redefine low-level vision.
The official LoViF @ CVPR 2026 workshop program will feature invited talks, oral paper presentations, poster discussions, challenge award announcements, and certificate information. ((poster session needs to wait the arrangement of workshop chairs)
Date: June 4th, 2026, AM | Venue: Room 504, Colorado Convention Center, Denver, USA | Time Zone: Mountain Time, MT
| Time | Session | Speaker / Presenter |
|---|---|---|
| 08:20 – 08:30 | Opening Remarks Welcome and introduction to LoViF 2026 | Workshop Organizers |
| 08:30 – 08:35 | Oral Presentation Session I π-Diff: Physically-Inspired Low-Light Image Enhancement with Structure Preserving Diffusion Priors | Recorded Videos by Min Wang, Northwestern Polytechnical University |
| 08:35 – 08:40 | Oral Presentation Session I PyrLight: Laplacian-Pyramid-Guided Conditional Diffusion for Detail-Preserving and Artifact-Free Low-Light Enhancement | Recorded Videos by Mingfu Yan, Southeast University |
| 08:40 – 08:50 | Oral Presentation Session I Controllable Image Generation with Composed Parallel Token Prediction | Jamie Stirling, Durham University |
| 08:50 – 09:00 | Oral Presentation Session I Don’t Waste Bits! Adaptive KV-Cache Quantization for Lightweight On-Device LLMs | Sayed Pedram Haeri, Clemson University |
| 09:00 – 09:30 | Invited Talk 1 Invited speaker talk | Prof. Dimitris Samaras Stony Brook University |
| 09:30 – 10:00 | Invited Talk 2 Invited speaker talk | Prof. Robby T. Tan NUS |
| 10:00 – 10:10 | Coffee Break Poster Session (you can have the poster between 9:30 AM and 12:00 AM. The length for it is not limited) | Poster Session |
| 10:10 – 10:20 | Oral Presentation Session II Towards Unified Image Deblurring using a Mixture-of-Experts Decoder | Marcos V. Conde, University of Würzburg & CIDAUT |
| 10:20 – 10:30 | Oral Presentation Session II Generative Panoramic Image Stitching | Mathieu Tuli, Meta |
| 10:30 – 11:00 | Invited Talk 3 Invited speaker talk | Dr. Yuyang Hu Research Scientist at Google |
| 11:00 – 11:30 | Invited Talk 4 Invited speaker talk | Prof. Michael S. Brown York University | 11:30 – 11:40 | Oral Presentation Session III EvoRender: An Agentic AI -- Neural Rendering Algorithm Discovery | Matai, Janarbek, AMD |
| 11:40 – 12:00 | Oral Presentation Session III Oral presentations 4–12, 5 minutes per presentation | Challenge Reports |
| 12:00 – 12:20 | Challenge Awards, Certificates & Closing Remarks Challenge award announcements, certificate information, and closing remarks | Workshop Organizers |
Each invited talk is scheduled for 30 minutes. Each oral presentation is scheduled for 6 minutes. The coffee break will be held together with the poster session. Challenge award announcements and certificate details will be updated after confirmation from challenge organizers and winning teams.
We are honored to invite leading researchers from academia and industry to share their insights at the workshop.
The following works have been accepted to LoViF @ CVPR 2026. Accepted papers and challenge reports are listed separately. Papers selected for oral presentation are marked with an “Oral Presentation” badge.
Papers selected for oral presentation are marked with an “Oral Presentation” badge. Challenge Reports are also Oral Presentation Other accepted papers and challenge reports will be presented during the accepted works poster session. The final presentation order and additional materials will be updated later.
The official award certificates for LoViF @ CVPR 2026 workshop challenges are available below. Participants may preview and download certificates by challenge category.
This preview displays only the certificates for the selected challenge category.
Award certificates for the Real-world All-in-One Image Restoration Challenge.
Award certificates for the Challenge on Weather Removal in Videos.
Award certificates for the Efficient VLM for Multimodal Ad Creative Quality Scoring Challenge.
Award certificates for the Human-oriented Semantic Image Quality Assessment Challenge.
Award certificates for the Holistic Quality Assessment for 4D World Model Challenge.
Award certificates for the Objective Track of the Short-form UGC Video Restoration Challenge (Objective Track).
We invite submissions on topics bridging low-level vision with AGI trends. The submission follows the template of CVPR2026 and does not exceed 8 pages (not including references)
Low-level vision (e.g. restoration, enhancement, compression, perception) with visual autoregressive modeling (VAR), diffusion models, and GAN...; Low-level vision for AI-generated contents (AIGC images/videos and others);...
Explainable quality assessment and reward models for generation or processing; Preference optimization (e.g., DPO, GRPO) for visual processing and perception; ...
Human-machine interaction theory in visual processing; Low-level agentic systems with LLMs/LVMs; Universal framework design; ...
Latent compression/Cache/Acceleration/Quantization/sparse attention for representation, visual signal generation or low-level processing; Low-level token compression and visual tokenizer; ...
Data coding, streaming, and visual processing for machine and intelligent tasks, such as large foundation models (LVMs), autonomous vehicles, robotics, or UAVs...
3D/4D (gaussian splatting, Nerf, point cloud...) compression, reconstruction/generation, audio-visual benchmark, editing, gaussian splatting, world model evaluation, rendering \& simulation pipelines...
Multi-modal low-level visual signal processing, compression and related instruction tuning, prompt techniques and zero/few-shot in-context learning...
Ethical and impacts of generative low-level vision, including privacy in agentic systems, safety and robustness for downstream tasks, and impacts in healthcare and biomedical imaging...
Participate in our challenge tracks.
Establishing a new benchmark for complex real-world degradations, this challenge evaluates all-in-one restoration through both quantitative metrics and visual quality.
Tasked with predicting product ratings from images, videos, and text, this challenge focuses on developing lightweight, high-performance VLMs for real-world creative quality assessment.
This competition aims to facilitate a reliable semantic quality assessment benchmark, which is more deeply aligned with human semantic understanding, intending to promote future research on semantics-oriented visual coding, transmission, and processing, .
This competition aims to introduce a comprehensive benchmark for evaluating the physical reasonableness of videos generated by world models (including 2D、3D and 4D generation models).
This challenge focuses on video restoration under complex weather conditions, emphasizing temporal consistency and robustness through the integration of physical priors and temporal reasoning.
OpenReview system opens for submissions
OpenReview opens
Challenge websites open for submission
Challenge tracks open
Workshop papers (11:59 PM AoE)
Deadline (AoE)
Decisions sent to authors
Decisions sent
USTC
Tencent
UT Austin
NTU
Snap Research
UT Austin
NUS
Kuaishou
USTC
USTC
NUS
Univ. of Trento
York Univ.
EntroVision
EntroVision
NTU
SYSU
NTU
SYSU
HFUT
Leeds
NTU
EMERGETECH
USTC
EIT
Tsinghua
Tsinghua
