Home Console Docs News Papers Tutorials Datasets Wiki SOTA LLM Models GPU Leaderboard Events

English

Image Inpainting On Places2 1

Metrics

FID

P-IDS

U-IDS

Results

Performance results of various models on this benchmark

				Paper Title	Repository
HFill	28.92	1.24	11.24	Contextual Residual Aggregation for Ultra High-Resolution Image Inpainting
TFill (20-50% free-form)	22.13	-	-	Bridging Global Context Interactions for High-Fidelity Image Completion
EdgeConnect	12.66	1.93	15.87	EdgeConnect: Generative Image Inpainting with Adversarial Edge Learning
AOT GAN	10.64	3.07	19.92	Aggregated Contextual Transformations for High-Resolution Image Inpainting
CR-Fill	9.657	5.53	22.90	CR-Fill: Generative Image Inpainting With Auxiliary Contextual Reconstruction	-
DeepFill v2	9.27	4.01	21.32	Free-Form Image Inpainting with Gated Convolution
ProFill	7.7	3.87	21.19	High-Resolution Image Inpainting with Iterative Confidence Feedback and Guided Upsampling
MADF	7.53	6.00	23.78	Image Inpainting by End-to-End Cascaded Refinement with Mask Awareness
LAMA	2.97	13.09	32.29	Resolution-robust Large Mask Inpainting with Fourier Convolutions
CoModGAN	2.92	19.64	35.78	Large Scale Image Completion via Co-Modulated Generative Adversarial Networks
MAT	1.96	23.42	38.34	MAT: Mask-Aware Transformer for Large Hole Image Inpainting
CM-GAN	1.628	20.96	37.42	CM-GAN: Image Inpainting with Cascaded Modulation GAN and Object-Aware Training
SEELE	1.477	23.40	39.00	Repositioning the Subject within Image
ASUKA	1.230	28.70	41.30	Towards Enhanced Image Inpainting: Mitigating Unwanted Object Insertion and Preserving Color Consistency

0 of 14 row(s) selected.

Image Inpainting On Places2 1 | SOTA | HyperAI