Nvidia Inpaint

Powered by NVIDIA CUDA, cuDNN, and NVIDIA Graphics, the video enhancer can speed up the conversion process beyond expectations. Meanwhile, you can enjoy this free video enhancing software within 30-days before your final subscription. Ee speed test. Use NVIDIA AI photo restoration technology to remove unwanted items in old photos; Cons: Limited feature to restore old photos; Also read our review to find out best image inpainting tools: Top 10 Inpaint Alternatives to Remove Object from Photo Review. Price: Free; Compatibility: Online.

Nvidia InpaintNvidia Inpaint
1CUHK - SenseTime Joint Lab, The Chinese University of Hong Kong
2Nanyang Technological University, Singapore

Download Demo
Nvidia inpaint

Abstract


Video inpainting, which aims at filling in missing regions of a video, remains challenging due to the difficulty of preserving the precise spatial and temporal coherence of video contents. In this work we propose a novel flow-guided video inpainting approach. Rather than filling in the RGB pixels of each frame directly, we consider video inpainting as a pixel propagation problem. We first synthesize a spatially and temporally coherent optical flow field across video frames using a newly designed Deep Flow Completion network. Then the synthesized flow field is used to guide the propagation of pixels to fill up the missing regions in the video. Specifically, the Deep Flow Completion network follows a coarse-to-fine refinement to complete the flow fields, while their quality is further improved by hard flow example mining. Following the guide of the completed flow, the missing video regions can be filled up precisely. Our method is evaluated on DAVIS and YouTube-VOS datasets qualitatively and quantitatively, achieving the state-of-the-art performance in terms of inpainting quality and speed.

Framework


Nvidia Inpainting

Our framework contains two steps, the first step is to complete the missing flow while the second step is to propagate pixels with the guidance of completed flow fields. In the first step, a Deep Flow Completion Network (DFC-Net) is proposed for coarse-to-fine flow completion. DFC-Net consists of three similar subnetworks named as DFC-S. The first subnetwork estimates the flow in a relatively coarse scale and feeds them into the second and third subnetwork for further refinement. In the second step, after the flow is obtained, most of the missing regions can be filled up by pixels in known regions through a flow-guided propagation from different frames. A conventional image inpainting network is finally employed to complete the remaining regions that are not seen in the entire video. Thanks to the high-quality estimated flow in the first step, we can easily propagate these image inpainting results to the entire video sequence.

Visualization


Results of our flow-guided video inpainting approach. For each input sequence (odd row), we show representative frames with mask of missing region overlay. We show the inpainting results in even rows. (Best viewed with zoom-in.)

Comparision with Huang et al.

Nvidia Image Inpainting


Citation

Celsys announced the results of the 27th International Illustration Contest, sponsored by NVIDIA Studio, the pioneer of GPU accelerated computing, today on April 28 (Wed).
27th International Illustration Contest Winning Entries
https://www.clipstudio.net/en/cspcontest27/
Celsys regularly holds illustration contests for users around the world. The 27th International Illustration Contest opened for submissions on March 16, with the theme My Time to Shine! Approximately 3,800 entries were gathered over the 23-day entry period from all around the world. Entries came from many countries and regions, created with multiple kinds of digital and traditional media. The winners will receive cash and prizes including an NVIDIA RTX 3070 GPU Graphics Card.
Winning Entries
Grand Prize
Artist: XiaoSA 小颯
Judges' Comments
What a wonderful scene of friends having fun drawing together! That mirror showing the face of the one drawing is very well placed - a great touch indeed. (Celsys)
Runner-up Prize
Artist: Hace Tres Iqus
Judges' Comments
This is a unique approach to expressing how you feel when you draw. The color choice and composition are top-notch - this is one of those works you never get tired of looking at. (Celsys)
Prizes
Grand Prize (1 winner)
US$1,500 cash prize, NVIDIA RTX 3070 GPU graphics card and CLIP STUDIO PAINT EX Activation Code (Dual-device plan, 3 years) or 10,000 GOLD
Runner-up Prize (1 winner)
US$1,000 cash prize and CLIP STUDIO PAINT EX Activation Code (Dual-device plan, 3 years) or 10,000 GOLD
3rd - 5th place (3 winners)
US$200 cash prize and CLIP STUDIO PAINT EX Activation Code (Dual-device plan, 1 year) or 5,000 GOLD
6th - 10th place (5 winners)
US$100 cash prize and CLIP STUDIO PAINT PRO Activation Code (Dual-device plan, 1 year) or 3,000 GOLD
Honorable mentions (10 winners)
CLIP STUDIO PAINT PRO Activation Code (Dual-device plan, 1 year) or 3,000 GOLD
27th International Illustration Contest
https://www.clipstudio.net/en/cspcontest27/
Sponsor
NVIDIA Studio
https://www.nvidia.com/en-us/studio/
NVIDIA Studio is a platform built to support all creators. NVIDIA Studio laptops and desktops are purpose-built to help you do your best work, fast.
Organizer
CELSYS,Inc.
https://www.clipstudio.net/en/
CELSYS,Inc.
Celsys will continue supporting creators to create artistic content with digital technology.
We provide solutions for content creation, distribution, and browsing, including support of creative activities through our illustration, manga and animation production software “CLIP STUDIO PAINT” and web service “Clip Studio”, as well as our e-book solution, “Clip Studio Reader.”
Corporate site: https://www.Celsys.co.jp/en/
CLIP STUDIO PAINT site: http://www.clipstudio.net/en/
Instagram: https://www.instagram.com/clipstudioofficial/
Twitter: https://twitter.com/clipstudiopaint
Facebook: https://www.Facebook.com/Celsys.clipstudiopaint/
YouTube: https://www.YouTube.com/c/CLIPSTUDIOPAINTchannel

Nvidia Image Inpainting Demo

discuss this in the forum

Online Inpaint

Press Release homepage / archives