Skip to content

Latest commit

 

History

History
executable file
·
727 lines (489 loc) · 42.1 KB

results.md

File metadata and controls

executable file
·
727 lines (489 loc) · 42.1 KB
Raw

Image classification task results

ResNet

Paper:https://arxiv.org/abs/1512.03385

DarkNet

Paper:https://arxiv.org/abs/1804.02767?e05802c1_page=1

RepVGG

Paper:https://arxiv.org/abs/2101.03697

RegNet

Paper:https://arxiv.org/abs/2003.13678

ViT

Paper:https://arxiv.org/abs/2010.11929

VAN

Paper:https://arxiv.org/abs/2202.09741

ResNetCifar training from scratch on CIFAR100

ResNetCifar is different from ResNet in the first few layers.

Network macs params input size gpu num batch epochs Top-1
ResNet18Cifar 557.935M 11.220M 32x32 1 RTX A5000 128 200 77.110
ResNet34Cifar 1.164G 21.328M 32x32 1 RTX A5000 128 200 78.140
ResNet50Cifar 1.312G 23.705M 32x32 1 RTX A5000 128 200 75.610
ResNet101Cifar 2.531G 42.697M 32x32 1 RTX A5000 128 200 76.970
ResNet152Cifar 3.751G 58.341M 32x32 1 RTX A5000 128 200 77.710

You can find more model training details in classification_training/cifar100/.

ResNet training from scratch on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
ResNet18 1.819G 11.690M 224x224 2 RTX A5000 256 100 70.512
ResNet34 3.671G 21.798M 224x224 2 RTX A5000 256 100 73.680
ResNet50 4.112G 25.557M 224x224 2 RTX A5000 256 100 76.300
ResNet101 7.834G 44.549M 224x224 2 RTX A5000 256 100 77.380
ResNet152 11.559G 60.193M 224x224 2 RTX A5000 256 100 77.542

You can find more model training details in classification_training/imagenet/.

DarkNet training from scratch on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
DarkNetTiny 412.537M 2.087M 256x256 2 RTX A5000 256 100 57.786
DarkNet19 3.663G 20.842M 256x256 2 RTX A5000 256 100 74.248
DarkNet53 9.322G 41.610M 256x256 2 RTX A5000 256 100 76.352

You can find more model training details in classification_training/imagenet/.

RepVGG training from scratch on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
RepVGG_A0_deploy 1.362G 8.309M 224x224 2 RTX A5000 256 120 72.010
RepVGG_A1_deploy 2.364G 12.790M 224x224 2 RTX A5000 256 120 74.032
RepVGG_A2_deploy 5.117G 25.500M 224x224 2 RTX A5000 256 120 76.078
RepVGG_B0_deploy 3.058G 14.339M 224x224 2 RTX A5000 256 120 74.880
RepVGG_B1_deploy 11.816G 51.829M 224x224 2 RTX A5000 256 120 77.790
RepVGG_B2_deploy 18.377G 80.315M 224x224 2 RTX A5000 256 120 78.120

You can find more model training details in classification_training/imagenet/.

RegNet training from scratch on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
RegNetX_400MF 410.266M 5.158M 224x224 2 RTX A5000 4096 300 69.466
RegNetX_600MF 616.813M 6.196M 224x224 2 RTX A5000 4096 300 71.754
RegNetX_800MF 820.324M 7.260M 224x224 2 RTX A5000 4096 300 73.148
RegNetX_1_6GF 1.635G 9.190M 224x224 2 RTX A5000 4096 300 76.142
RegNetX_3_2GF 3.222G 15.297M 224x224 2 RTX A5000 4096 300 78.244
RegNetX_4_0GF 4.013G 22.118M 224x224 2 RTX A5000 4096 300 78.916

You can find more model training details in classification_training/imagenet/.

U2NetBackbone training from scratch on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
U2NetBackbone 13.097G 26.181M 224x224 2 RTX A5000 256 100 76.038

You can find more model training details in classification_training/imagenet/.

ResNet finetune from ImageNet21k pretrain weight on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
ResNet18 1.819G 11.690M 224x224 2 RTX A5000 4096 300 71.580
ResNet34 3.671G 21.798M 224x224 2 RTX A5000 4096 300 76.316
ResNet50 4.112G 25.557M 224x224 2 RTX A5000 4096 300 79.484
ResNet101 7.834G 44.549M 224x224 2 RTX A5000 4096 300 80.940
ResNet152 11.559G 60.193M 224x224 2 RTX A5000 4096 300 81.236

You can find more model training details in classification_training/imagenet/.

ResNet finetune from DINO pretrain weight on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
ResNet18 1.819G 11.690M 224x224 1 RTX A5000 256 100 70.754
ResNet18 1.819G 11.690M 224x224 1 RTX A5000 4096 300 71.362
ResNet34 3.671G 21.798M 224x224 2 RTX A5000 256 100 74.218
ResNet34 3.671G 21.798M 224x224 2 RTX A5000 4096 300 75.916
ResNet50 4.112G 25.557M 224x224 2 RTX A5000 256 100 77.114
ResNet50 4.112G 25.557M 224x224 2 RTX A5000 4096 300 79.418

You can find more model training details in classification_training/imagenet/.

ViT finetune from self-trained MAE pretrain weight(400epoch) on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
ViT-Tiny-Patch16 1.075G 5.670M 224x224 1 RTX A5000 4096 100 68.614
ViT-Small-Patch16 4.241G 21.955M 224x224 2 RTX A5000 4096 100 79.006
ViT-Base-Patch16 16.849G 86.377M 224x224 2 RTX A5000 4096 100 83.204
ViT-Large-Patch16 59.647G 304.024M 224x224 2 RTX A5000 4096 100 85.020

You can find more model training details in classification_training/imagenet/.

ViT finetune from offical MAE pretrain weight(800 epoch) on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
ViT-Base-Patch16 16.849G 86.377M 224x224 2 RTX A5000 4096 100 83.290
ViT-Large-Patch16 59.647G 304.024M 224x224 2 RTX A5000 4096 100 85.876

You can find more model training details in classification_training/imagenet/.

ResNet train from ImageNet1K pretrain weight on ImageNet21K(Winter 2021 release)

Network macs params input size gpu num batch epochs Semantic Softmax Acc
ResNet18 1.819G 11.690M 224x224 2 RTX A5000 4096 80 68.639
ResNet34 3.671G 21.798M 224x224 2 RTX A5000 4096 80 71.873
ResNet50 4.112G 25.557M 224x224 2 RTX A5000 4096 80 74.664
ResNet101 7.834G 44.549M 224x224 2 RTX A5000 4096 80 76.136
ResNet152 11.559G 60.193M 224x224 2 RTX A5000 4096 80 75.731

You can find more model training details in classification_training/imagenet21k/.

ViT finetune from self-trained MAE pretrain weight(100epoch) on ACCV2022

Network macs params input size gpu num batch epochs Top-1
ViT-Large-Patch16 59.651G 308.124M 224x224 2 RTX 4090 4096 100 90.693

You can find more model training details in classification_training/accv2022/.

VAN finetune from offical pretrain weight on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
VAN-B0 880.224M 4.103M 224x224 2 RTX A5000 1024 300 75.618
VAN-B1 2.518G 13.856M 224x224 2 RTX 4090 1024 300 80.956
VAN-B2 5.033G 26.567M 224x224 4 RTX 4090 1024 300 82.322

You can find more model training details in classification_training/imagenet/.

Object detection task results

RetinaNet

Paper:https://arxiv.org/abs/1708.02002

FCOS

Paper:https://arxiv.org/abs/1904.01355

CenterNet

Paper:https://arxiv.org/abs/1904.07850

TTFNet

Paper:https://arxiv.org/abs/1909.00700

DETR

Paper:https://arxiv.org/abs/2005.12872

DINO-DETR

Paper:https://arxiv.org/abs/2203.03605

All detection models training from scratch on COCO2017

Trained on COCO2017 train dataset, tested on COCO2017 val dataset.

mAP is IoU=0.5:0.95,area=all,maxDets=100,mAP(COCOeval,stats[0]).

Network resize-style input size macs params gpu num batch epochs mAP
ResNet50-RetinaNet YoloStyle-640 640x640 95.558G 37.969M 2 RTX A5000 32 13 34.459
ResNet50-RetinaNet YoloStyle-800 800x800 149.522G 37.969M 2 RTX A5000 32 13 36.023
ResNet50-RetinaNet RetinaStyle-800 800x1333 250.069G 37.969M 2 RTX A5000 8 13 35.434
ResNet50-FCOS YoloStyle-640 640x640 81.943G 32.291M 2 RTX A5000 32 13 37.176
ResNet50-FCOS YoloStyle-800 800x800 128.160G 32.291M 2 RTX A5000 32 13 38.745
ResNet50-FCOS RetinaStyle-800 800x1333 214.406G 32.291M 2 RTX A5000 8 13 39.649
ResNet18DCN-CenterNet YoloStyle-512 512x512 14.854G 12.889M 2 RTX A5000 64 140 26.209
ResNet18DCN-TTFNet-3x YoloStyle-512 512x512 16.063G 13.737M 2 RTX A5000 64 39 27.054
ResNet50-DETR YoloStyle-1024 1024x1024 89.577G 30.440M 8 RTX A5000 64 500 36.941
ResNet50-DINO-DETR YoloStyle-1024 1024x1024 844.204G 47.082M 8 RTX A5000 16 13 42.870
ResNet50-DINO-DETR YoloStyle-1024 1024x1024 844.204G 47.082M 8 RTX A5000 16 39 45.445

You can find more model training details in detection_training/coco/.

All detection models finetune from objects365 pretrain weight on COCO2017

Trained on COCO2017 train dataset, tested on COCO2017 val dataset.

mAP is IoU=0.5:0.95,area=all,maxDets=100,mAP(COCOeval,stats[0]).

Network resize-style input size macs params gpu num batch epochs mAP
ResNet50-RetinaNet YoloStyle-640 640x640 95.558G 37.969M 2 RTX A5000 32 13 38.930
ResNet50-RetinaNet YoloStyle-800 800x800 149.522G 37.969M 2 RTX A5000 32 13 40.483
ResNet50-RetinaNet RetinaStyle-800 800x1333 250.069G 37.969M 2 RTX A5000 8 13 40.424
ResNet50-FCOS YoloStyle-640 640x640 81.943G 32.291M 2 RTX A5000 32 13 42.871
ResNet50-FCOS YoloStyle-800 800x800 128.160G 32.291M 2 RTX A5000 32 13 44.526
ResNet50-FCOS RetinaStyle-800 800x1333 214.406G 32.291M 2 RTX A5000 8 13 42.848

You can find more model training details in detection_training/coco/.

All detection models train from COCO2017 pretrain weight on Objects365(v2,2020)

Trained on objects365 train dataset, tested on objects365 val dataset.

mAP is IoU=0.5:0.95,area=all,maxDets=100,mAP(COCOeval,stats[0]).

Network resize-style input size macs params gpu num batch epochs mAP
ResNet50-RetinaNet YoloStyle-800 800x800 149.522G 37.969M 8 RTX A5000 32 13 16.360
ResNet50-FCOS RetinaStyle-800 800x1333 214.406G 32.291M 8 RTX A5000 32 13 17.068

All detection models training from scratch on VOC2007 and VOC2012

Trained on VOC2007 trainval dataset + VOC2012 trainval dataset, tested on VOC2007 test dataset.

mAP is IoU=0.50,area=all,maxDets=100,mAP.

Network resize-style input size macs params gpu num batch epochs mAP
ResNet50-RetinaNet YoloStyle-640 640x640 84.947G 36.724M 2 RTX A5000 32 13 81.948
ResNet50-FCOS YoloStyle-640 640x640 80.764G 32.153M 2 RTX A5000 32 13 81.624

You can find more model training details in detection_training/voc/.

All detection models finetune from objects365 pretrain weight on VOC2007 and VOC2012

Trained on VOC2007 trainval dataset + VOC2012 trainval dataset, tested on VOC2007 test dataset.

mAP is IoU=0.50,area=all,maxDets=100,mAP.

Network resize-style input size macs params gpu num batch epochs mAP
ResNet50-RetinaNet YoloStyle-640 640x640 84.947G 36.724M 2 RTX A5000 32 13 90.220
ResNet50-FCOS YoloStyle-640 640x640 80.764G 32.153M 2 RTX A5000 32 13 90.371

You can find more model training details in detection_training/voc/.

Semantic Segmentation task results

DeepLabv3+

Paper:https://arxiv.org/abs/1802.02611

U2Net

Paper:https://arxiv.org/abs/2005.09007

All semantic segmentation models training from scratch on ADE20K

Network input size macs params gpu num batch epochs miou
ResNet50-DeepLabv3+ 512x512 25.548G 26.738M 2 RTX A5000 8 128 34.659
U2Net 512x512 219.012G 46.191M 2 RTX A5000 8 128 39.046

You can find more model training details in semantic_segmentation_training/ade20k/.

All semantic segmentation models training from scratch on COCO2017

Network input size macs params gpu num batch epochs miou
ResNet50-DeepLabv3+ 512x512 25.548G 26.738M 2 RTX A5000 32 64 64.176
U2Net 512x512 219.012G 46.191M 4 RTX A5000 32 64 66.529

You can find more model training details in semantic_segmentation_training/coco/.

Instance Segmentation task results

YOLACT

Paper:https://arxiv.org/abs/1904.02689

SOLOv2

Paper:https://arxiv.org/abs/2003.10152

All instance segmentation models training from scratch on COCO2017

Trained on COCO2017 train dataset, tested on COCO2017 val dataset.

mAP is IoU=0.5:0.95,area=all,maxDets=100,mAP(COCOeval,stats[0]).

Network resize-style input size macs params gpu num batch epochs mAP
ResNet50-YOLACT YoloStyle-800 800x800 123.095G 31.165M 4 RTX A5000 64 39 28.061
ResNet50-SOLOv2 YoloStyle-1024 1024x1024 248.546G 46.582M 4 RTX A5000 32 39 36.559

You can find more model training details in instance_segmentation_training/coco/.

Knowledge distillation task results

KD loss

Paper:https://arxiv.org/abs/1503.02531

DML loss

Paper:https://arxiv.org/abs/1706.00384

ResNet training from pretrain weight on ImageNet1K(ILSVRC2012)

Teacher Network Student Network method Freeze Teacher input size gpu num batch epochs Teacher Top-1 Student Top-1
ResNet152 ResNet50 CE+KD True 224x224 2 RTX A5000 256 100 / 77.352
ResNet152 ResNet50 CE+DML False 224x224 2 RTX A5000 256 100 79.274 78.122
ResNet152 ResNet50 CE+KD+Vit Aug True 224x224 2 RTX A5000 4096 300 / 80.168
ResNet152 ResNet50 CE+DML+Vit Aug False 224x224 2 RTX A5000 4096 300 81.508 79.810

You can find more model training details in distillation_training/imagenet/.

Contrastive learning task results

DINO:Emerging Properties in Self-Supervised Vision Transformers

Paper:https://arxiv.org/abs/2104.14294

ResNet DINO pretrain on ImageNet1K(ILSVRC2012)

Network input size gpu num batch epochs Loss
ResNet18-DINO 224x224 4 RTX A5000 256 400 3.081
ResNet34-DINO 224x224 4 RTX A5000 256 400 2.425
ResNet50-DINO 224x224 4 RTX A5000 256 400 1.997

You can find more model training details in contrastive_learning_training/imagenet/.

ResNet finetune from DINO pretrain weight on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
ResNet50 4.112G 25.557M 224x224 2 RTX A5000 256 100 77.114
ResNet50 4.112G 25.557M 224x224 2 RTX A5000 4096 300 79.418

You can find more model training details in classification_training/imagenet/.

Masked image modeling task results

MAE:Masked Autoencoders Are Scalable Vision Learners

Paper:https://arxiv.org/abs/2111.06377

ViT MAE pretrain on ImageNet1K(ILSVRC2012)

Network input size gpu num batch epochs Loss
ViT-Tiny-Patch16 224x224 1 RTX A5000 256 400 0.427
ViT-Small-Patch16 224x224 2 RTX A5000 256 400 0.414
ViT-Base-Patch16 224x224 2 RTX A5000 256 400 0.388
ViT-Large-Patch16 224x224 2 RTX A5000 256 400 0.378

You can find more model training details in masked_image_modeling_training/imagenet/.

ViT MAE pretrain on ACCV2022 from ImageNet1K pretrain

Network input size gpu num batch epochs Loss
ViT-Large-Patch16 224x224 2 RTX 4090 256 100 0.423

You can find more model training details in masked_image_modeling_training/accv2022/.

ViT finetune from self-trained MAE pretrain weight(400epoch) on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
ViT-Tiny-Patch16 1.075G 5.670M 224x224 1 RTX A5000 4096 100 68.614
ViT-Small-Patch16 4.241G 21.955M 224x224 2 RTX A5000 4096 100 79.006
ViT-Base-Patch16 16.849G 86.377M 224x224 2 RTX A5000 4096 100 83.204
ViT-Large-Patch16 59.647G 304.024M 224x224 2 RTX A5000 4096 100 85.020

You can find more model training details in classification_training/imagenet/.

ViT finetune from offical MAE pretrain weight(800 epoch) on ImageNet1K(ILSVRC2012)

Network macs params input size gpu num batch epochs Top-1
ViT-Base-Patch16 16.849G 86.377M 224x224 2 RTX A5000 4096 100 83.290
ViT-Large-Patch16 59.647G 304.024M 224x224 2 RTX A5000 4096 100 85.876

You can find more model training details in classification_training/imagenet/.

ViT finetune from self-trained MAE pretrain weight(100epoch) on ACCV2022

Network macs params input size gpu num batch epochs Top-1
ViT-Large-Patch16 59.651G 308.124M 224x224 2 RTX 4090 4096 100 90.693

You can find more model training details in classification_training/accv2022/.

OCR text detection task results

DBNet

Paper:https://arxiv.org/abs/1911.08947

Use combine dataset include ICDAR2017RCTW/ICDAR2019ART/ICDAR2019LSVT/ICDAR2019MLT to train and test.

Network macs params input size gpu num batch epochs precision recall f1
repvgg_dbnet 11.806G 726.338K 960x960 2 RTX A5000 128 200 88.756 74.205 80.831
resnet50_dbnet 139.141G 24.784M 960x960 2 RTX A5000 64 100 92.973 86.316 89.521
vanb1_dbnet 108.596G 14.439M 960x960 2 RTX A5000 64 100 93.049 86.881 89.859

You can find more model training details in ocr_text_detection_training/.

OCR text recognition task results

CRNN+LSTM+CTC

Paper:https://arxiv.org/abs/1507.05717

Use combine dataset aistudio_baidu_street/chinese_dataset/synthetic_chinese_string_dataset/meta_self_learning_dataset to train and test.

Network macs params input size gpu num batch epochs lcs_precision lcs_recall
repvgg_ctc_model 951.804M 6.865M 32x512 2 RTX A5000 512 50 98.079 97.564
resnet50_ctc_model 12.474G 179.864M 32x512 2 RTX A5000 1024 50 99.368 99.135
van_b1_ctc_model 2.410G 27.954 M 32x512 2 RTX A5000 1024 50 98.868 97.597

You can find more model training details in ocr_text_recognition_training/.

Human matting task results

PFAN+Matting

Paper1:https://arxiv.org/abs/1903.00179

Paper2:https://arxiv.org/abs/2104.14222

Paper3:https://arxiv.org/abs/2202.09741

Use combine dataset Deep_Automatic_Portrait_Matting/RealWorldPortrait636/P3M10K to train and test.

Network macs params input size gpu num batch epochs iou precision recall sad mae mse grad conn
resnet50_pfan_matting 85.638G 29.654M 832x832 2 RTX 3090 32 50 0.9818 0.9879 0.9937 5.9215 0.0085 0.0048 7.5277 5.6842
van_b2_pfan_matting 85.926G 27.854M 832x832 2 RTX 3090 32 50 0.9850 0.9900 0.9948 5.0200 0.0072 0.0038 5.5563 4.7644

You can find more model training details in human_matting_training/.

Salient object detection task results

PFAN+Segmentation

Paper1:https://arxiv.org/abs/1903.00179

Paper2:https://arxiv.org/abs/2202.09741

Use combine dataset DIS5K/HRS10K/HRSOD/UHRSD to train and test.

Network macs params input size gpu num batch epochs iou precision recall f_squared_beta
resnet50_pfan_segmentation 70.921G 26.580M 832x832 8 RTX 4090D 64 100 0.8501 0.8977 0.9389 0.9068
van_b2_pfan_segmentation 77.433G 26.953M 832x832 8 RTX 4090D 64 100 0.8904 0.9292 0.9527 0.9345

You can find more model training details in salient_object_detection_training/.

Face detection task results

RetinaFace

Paper:https://arxiv.org/pdf/1905.00641

Use WiderFace train and UFDD val datasets to train, WiderFace val dataset to test.

Network macs params input size gpu num batch epochs decode setting Easy AP Medium AP Hard AP
resnet50_retinaface 100.372G 27.277M 960x960 2 RTX A5000 16 100 max_object_num:200
min_score_threshold:0.3
topn:1000		 0.9311 0.9043 0.7357
resnet50_retinaface 100.372G 27.277M 960x960 2 RTX A5000 16 100 max_object_num:1000
min_score_threshold:0.1
topn:2000		 0.9357 0.9158 0.8105

You can find more model training details in face_detection_training/.

Interactive segmentation task results

SAM(segment-anything)

Paper:https://arxiv.org/pdf/2304.02643

using random noise prompt box to test model.

Network dataset input size gpu num batch epochs iou precision recall
sam_b salient_object_detection 1024x1024 8 RTX 4090D 16 500 0.9486 0.9676 0.9783
sam_b sobav2 1024x1024 2 RTX 3090 4 500 0.9871 0.9935 0.9930
sam_b desobav2 1024x1024 8 RTX 4090D 16 200 0.9862 0.9919 0.9941

You can find more model training details in interactive_segmentation_training/.

Image inpainting model task results

Aggregated Contextual Transformations for High-Resolution Image Inpainting

Paper:https://arxiv.org/abs/2104.01431

All image inpainting model training from scratch on CelebA-HQ

Trained image inpainting model on CelebA-HQ dataset.Test image num=2000.

Network input size epochs Mask mae psnr ssim fid
AOT-GAN 512x512 100 0.01-0.1 0.0023 40.368 0.9853 0.8003
AOT-GAN 512x512 100 0.1-0.2 0.0064 33.724 0.9592 2.1704
AOT-GAN 512x512 100 0.2-0.3 0.0122 29.996 0.9245 3.8093
AOT-GAN 512x512 100 0.3-0.4 0.0192 27.343 0.8860 5.4981
AOT-GAN 512x512 100 0.4-0.5 0.0279 25.154 0.8426 8.3303
AOT-GAN 512x512 100 0.5-0.6 0.0486 21.576 0.7704 14.553

You can find more model training details in image_inpainting_training/celebahq/.

All image inpainting model training from scratch on Places365-standard

Trained image inpainting model on Places365-standard dataset.Test image num=36500.

Network input size epochs Mask mae psnr ssim fid
AOT-GAN 512x512 5 0.01-0.1 0.0041 35.505 0.9772 0.1412
AOT-GAN 512x512 5 0.1-0.2 0.0114 29.250 0.9374 0.4833
AOT-GAN 512x512 5 0.2-0.3 0.0214 25.802 0.8855 1.1973
AOT-GAN 512x512 5 0.3-0.4 0.0331 23.391 0.8291 2.5272
AOT-GAN 512x512 5 0.4-0.5 0.0469 21.504 0.7677 5.0670
AOT-GAN 512x512 5 0.5-0.6 0.0737 18.904 0.6795 14.951
Network input size epochs Mask mae psnr ssim fid
AOT-GAN-light 512x512 5 0.01-0.1 0.0043 35.023 0.9757 0.1680
AOT-GAN-light 512x512 5 0.1-0.2 0.0121 28.824 0.9338 0.6524
AOT-GAN-light 512x512 5 0.2-0.3 0.0227 25.423 0.8798 1.7831
AOT-GAN-light 512x512 5 0.3-0.4 0.0350 23.052 0.8218 4.0379
AOT-GAN-light 512x512 5 0.4-0.5 0.0494 21.199 0.7590 8.2494
AOT-GAN-light 512x512 5 0.5-0.6 0.0768 18.690 0.6719 22.745
TRANSX-LKA-AOT-GAN-light 512x512 1 0.01-0.1 0.0042 35.287 0.9763 0.1611
TRANSX-LKA-AOT-GAN-light 512x512 1 0.1-0.2 0.0117 29.148 0.9356 0.5648
TRANSX-LKA-AOT-GAN-light 512x512 1 0.2-0.3 0.0217 25.790 0.8835 1.3744
TRANSX-LKA-AOT-GAN-light 512x512 1 0.3-0.4 0.0331 23.451 0.8281 2.7966
TRANSX-LKA-AOT-GAN-light 512x512 1 0.4-0.5 0.0464 21.599 0.7681 5.1937
TRANSX-LKA-AOT-GAN-light 512x512 1 0.5-0.6 0.0718 19.034 0.6838 12.825

You can find more model training details in image_inpainting_training/places365_standard.

All image inpainting model training from scratch on Places365-challenge

Trained image inpainting model on Places365-challenge dataset.Test image num=36500.

Network input size epochs Mask mae psnr ssim fid
AOT-GAN 512x512 1 0.01-0.1 0.0039 35.807 0.9781 0.1318
AOT-GAN 512x512 1 0.1-0.2 0.0110 29.499 0.9395 0.4493
AOT-GAN 512x512 1 0.2-0.3 0.0207 26.021 0.8890 1.0881
AOT-GAN 512x512 1 0.3-0.4 0.0320 23.586 0.8338 2.2785
AOT-GAN 512x512 1 0.4-0.5 0.0454 21.674 0.7734 4.4948
AOT-GAN 512x512 1 0.5-0.6 0.0715 19.039 0.6848 13.475
Network input size epochs Mask mae psnr ssim fid
AOT-GAN-light 512x512 1 0.01-0.1 0.0042 35.263 0.9762 0.1609
AOT-GAN-light 512x512 1 0.1-0.2 0.0118 29.043 0.9349 0.6028
AOT-GAN-light 512x512 1 0.2-0.3 0.0221 25.609 0.8814 1.6013
AOT-GAN-light 512x512 1 0.3-0.4 0.0340 23.209 0.8235 3.5484
AOT-GAN-light 512x512 1 0.4-0.5 0.0480 21.332 0.7606 7.2095
AOT-GAN-light 512x512 1 0.5-0.6 0.0745 18.778 0.6714 20.031
TRANSX-LKA-AOT-GAN-light 512x512 1 0.01-0.1 0.0041 35.466 0.9769 0.1392
TRANSX-LKA-AOT-GAN-light 512x512 1 0.1-0.2 0.0114 29.275 0.9369 0.4566
TRANSX-LKA-AOT-GAN-light 512x512 1 0.2-0.3 0.0212 25.888 0.8854 1.0463
TRANSX-LKA-AOT-GAN-light 512x512 1 0.3-0.4 0.0325 23.530 0.8300 2.0431
TRANSX-LKA-AOT-GAN-light 512x512 1 0.4-0.5 0.0457 21.667 0.7699 3.8222
TRANSX-LKA-AOT-GAN-light 512x512 1 0.5-0.6 0.0711 19.104 0.6836 9.9868

You can find more model training details in image_inpainting_training/places365_challenge.

Diffusion model task results

Denoising Diffusion Probabilistic Models

Paper:https://arxiv.org/abs/2006.11239

Denoising Diffusion Implicit Models

Paper:https://arxiv.org/abs/2010.02502

High-Resolution Image Synthesis with Latent Diffusion Models

Paper:https://arxiv.org/abs/2112.10752

All diffusion model with different sampling methods on CIFAR10

Trained diffusion unet on CIFAR10 dataset(DDPM method).Test image num=50000.

sampling method input size steps condition label(train/test) FID IS score(mean/std)
DDPM 32x32 1000 False/False 5.394 8.684/0.169
DDIM 32x32 50 False/False 7.644 8.642/0.129
DDPM 32x32 1000 True/True 3.949 8.985/0.139

You can find more model training details in diffusion_model_training/cifar10/.

All diffusion model with different sampling methods on CIFAR100

Trained diffusion unet on CIFAR100 dataset(DDPM method).Test image num=50000.

sampling method input size steps condition label(train/test) FID IS score(mean/std)
DDPM 32x32 1000 False/False 9.620 9.399/0.138
DDIM 32x32 50 False/False 13.250 8.946/0.150
DDPM 32x32 1000 True/True 5.209 10.880/0.180

You can find more model training details in diffusion_model_training/cifar100/.

All diffusion model with different sampling methods on CelebA-HQ

Trained diffusion unet on CelebA-HQ dataset(DDPM method).Test image num=28000.

sampling method input size steps condition label(train/test) FID IS score(mean/std)
DDPM 64x64 1000 False/False 6.491 2.577/0.035
DDIM 64x64 50 False/False 15.195 2.625/0.028

You can find more model training details in diffusion_model_training/celebahq/.

All diffusion model with different sampling methods on FFHQ

Trained diffusion unet on FFHQ dataset(DDPM method).Test image num=60000.

sampling method input size steps condition label(train/test) FID IS score(mean/std)
DDPM 64x64 1000 False/False 6.671 3.399/0.055
DDIM 64x64 50 False/False 10.479 3.431/0.044

You can find more model training details in diffusion_model_training/ffhq/.