Adobe Research creates new technologies in computer vision and imaging, including breakthroughs in image search, organization, editing, and generation. Our scientists and engineers tackle problems such as object, text, and face detection, segmentation, person recognition, image quality, captioning, denoising, and deblurring. Many of our approaches tap into deep learning to empower creative image editing.
We also focus on video enhancement, analysis, synthesis, and understanding. We are interested in research problems such as video object segmentation, stabilization, generation, recognition, joint audio and video understanding, compression, and delivery. We seek to empower our users to create compelling video content and help our customers to maximize the value of their videos through next-generation experiences, analytics, and data-driven enhancement to video creation and monetization.
Meet some of our researchersView More
View our latest publicationsView More
Power Efficient Mobile VTuber Live Streaming
Petrangeli, S., Swaminathan, V. (Dec. 6, 2025)
ACM International Conference on Multimedia in Asia (MM Asia)
The Photographer’s Eye: Teaching Multimodal Large Language Models to See, and Critique Like Photographers
Qi, D., Zhao, H., Shi, J., Jenni, S., Fan, Y., Dernoncourt, F., Cohen, S., Li, S. (Jun. 15, 2025)
CVPR 2025
TIDE: Training Locally Interpretable Domain Generalization Models Enables Test-time Correction
Agarwal, A., Karanam, S., Gandhi, V. (Jun. 10, 2025)
Computer Vision and Pattern Recognition (CVPR)
Project Perfect Blend
Project Perfect Blend is a generative harmonization tool that makes compositing easier. When adding people or objects into another image, it will adjust color, lighting, shadows, and reflections to blend the foreground and background layers naturally. This project focuses on natural blending, foreground relighting, and realistic shadow casting to make the process smoother and more lifelike.
