SOTAVerified

Image Matting

Image Matting is the process of accurately estimating the foreground object in images and videos. It is a very important technique in image and video editing applications, particularly in film production for creating visual effects. In case of image segmentation, we segment the image into foreground and background by labeling the pixels. Image segmentation generates a binary image, in which a pixel either belongs to foreground or background. However, Image Matting is different from the image segmentation, wherein some pixels may belong to foreground as well as background, such pixels are called partial or mixed pixels. In order to fully separate the foreground from the background in an image, accurate estimation of the alpha values for partial or mixed pixels is necessary.

Source: Automatic Trimap Generation for Image Matting

Image Source: Real-Time High-Resolution Background Matting

Papers

Showing 76100 of 225 papers

TitleStatusHype
FADE: A Task-Agnostic Upsampling Operator for Encoder-Decoder ArchitecturesCode1
FADE: Fusing the Assets of Decoder and Encoder for Task-Agnostic UpsamplingCode1
Towards Label-Efficient Human Matting: A Simple Baseline for Weakly Semi-Supervised Trimap-Free Human MattingCode0
TOM-Net: Learning Transparent Object Matting from a Single ImageCode0
AlphaGAN: Generative adversarial networks for natural image mattingCode0
DFIMat: Decoupled Flexible Interactive Matting in Multi-Person ScenariosCode0
Information-Flow MattingCode0
Semantic Human MattingCode0
Sub-frame Appearance and 6D Pose Estimation of Fast Moving ObjectsCode0
Towards Real-Time Automatic Portrait Matting on Mobile DevicesCode0
AugStatic - A Light-Weight Image Augmentation LibraryCode0
Augmented Balanced Image Dataset Generator Using AugStatic LibraryCode0
A Late Fusion CNN for Digital MattingCode0
Intra-frame Object Tracking by DeblattingCode0
Attention Transfer Network for Nature Image MattingCode0
ReFlow6D: Refraction-Guided Transparent Object 6D Pose Estimation via Intermediate Representation LearningCode0
Deep-Energy: Unsupervised Training of Deep Neural NetworksCode0
PP-Matting: High-Accuracy Natural Image MattingCode0
Guided Collaborative Training for Pixel-wise Semi-Supervised LearningCode0
GIMP-ML: Python Plugins for using Computer Vision Models in GIMPCode0
Non-Causal Tracking by DeblattingCode0
PLADE-Net: Towards Pixel-Level Accuracy for Self-Supervised Single-View Depth Estimation with Neural Positional Encoding and Distilled Matting LossCode0
Real-time deep hair matting on mobile devicesCode0
Fourier-Domain Optimization for Image ProcessingCode0
MODNet-V: Improving Portrait Video Matting via Background RestorationCode0
Show:102550
← PrevPage 4 of 9Next →

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1DIMMSE14Unverified
2IndexNet-MattingMSE13Unverified
3Context-Aware MattingMSE8.2Unverified
4SIMMSE5.8Unverified
5LSAMattingMSE5.4Unverified
6FBAMattingMSE5.3Unverified
7PP-MattingMSE5Unverified
8LFPNetMSE4.1Unverified
9MatteFormerMSE4Unverified
10TMFNetMSE3.6Unverified
#ModelMetricClaimedVerifiedStatus
1SHMCSAD61.5Unverified
2LFSAD36.12Unverified
3HATTSAD28.01Unverified
4SHMSAD17.81Unverified
5GFM(r)SAD10.89Unverified
6GFM(d)SAD10.26Unverified
7GFM(r2b)SAD10.24Unverified
8GFM(r')SAD9.66Unverified
9StyleMatteSAD9.6Unverified
#ModelMetricClaimedVerifiedStatus
1LFSAD42.95Unverified
2HATTSAD25.99Unverified
3SHMSAD21.56Unverified
4GFMSAD13.2Unverified
5P3M-Net (r)SAD8.73Unverified
6StyleMatteSAD6.97Unverified
7P3M-Net (v)SAD6.24Unverified
#ModelMetricClaimedVerifiedStatus
1LFSAD191.74Unverified
2SHMSAD170.44Unverified
3U2NETSAD83.46Unverified
4GFMSAD52.66Unverified
5AIM-NetSAD43.92Unverified
6DiffMatteSAD16.31Unverified
#ModelMetricClaimedVerifiedStatus
1CAMMSE4.5Unverified
2BMMSE1.33Unverified
3IMMSE1.16Unverified
4Adobe LS-GANMSE0.97Unverified
#ModelMetricClaimedVerifiedStatus
1PP-MattingSAD40.69Unverified
2DCAMSAD31.27Unverified
3ViTMatteSAD17.05Unverified
4DiffMatteSAD15.5Unverified
#ModelMetricClaimedVerifiedStatus
1MODNet+MAD0.81Unverified
#ModelMetricClaimedVerifiedStatus
1MODNet+ (Our)MAD0.97Unverified