MaskFormer:每像素分类并不是语义分割所需要的全部
需积分: 35 91 浏览量
2021-08-05
00:58:46
上传
评论
收藏 348KB ZIP 举报
MaskFormer-master.zip （88个子文件） 
MaskFormer-master configs cityscapes-19 
maskformer_R101_bs16_90k.yaml 883B maskformer_R101c_bs16_90k.yaml 446B Base-Cityscapes-19.yaml 1KB ade20k-150-panoptic maskformer_panoptic_R50_bs16_720k.yaml 890B maskformer_panoptic_R101_bs16_720k.yaml 304B ade20k-full-847 Base-ADE20KFull-847.yaml 1KB maskformer_R50_bs16_200k.yaml 637B per_pixel_baseline_R50_bs16_200k.yaml 343B maskformer_R101c_bs16_200k.yaml 349B maskformer_R101_bs16_200k.yaml 295B per_pixel_baseline_plus_R50_bs16_200k.yaml 624B ade20k-150 Base-ADE20K-150.yaml 1KB maskformer_R101_bs16_160k.yaml 295B swin maskformer_swin_small_bs16_160k.yaml 547B maskformer_swin_base_IN21k_384_bs16_160k_res640.yaml 1KB maskformer_swin_tiny_bs16_160k.yaml 545B maskformer_swin_large_IN21k_384_bs16_160k_res640.yaml 1KB maskformer_R101c_bs16_160k.yaml 349B per_pixel_baseline_plus_R50_bs16_160k.yaml 618B per_pixel_baseline_R50_bs16_160k.yaml 337B maskformer_R50_bs16_160k.yaml 631B coco-stuff-10k-171 maskformer_R101_bs32_60k.yaml 294B maskformer_R101c_bs32_60k.yaml 348B per_pixel_baseline_plus_R50_bs32_60k.yaml 624B Base-COCOStuff10K-171.yaml 1KB maskformer_R50_bs32_60k.yaml 637B per_pixel_baseline_R50_bs32_60k.yaml 343B mapillary-vistas-65 maskformer_R50_bs16_300k.yaml 637B Base-MapillaryVistas-65.yaml 1KB coco-panoptic maskformer_panoptic_R50_bs64_554k.yaml 945B Base-COCO-PanopticSegmentation.yaml 1KB maskformer_panoptic_R101_bs64_554k.yaml 304B swin maskformer_panoptic_swin_large_IN21k_384_bs64_554k.yaml 1019B maskformer_panoptic_swin_base_IN21k_384_bs64_554k.yaml 822B maskformer_panoptic_swin_small_bs64_554k.yaml 790B maskformer_panoptic_swin_tiny_bs64_554k.yaml 788B GETTING_STARTED.md 2KB train_net.py 10KB
requirements.txt 30B demo demo.py 7KB README.md 163B predictor.py 8KB datasets prepare_ade20k_sem_seg.py 896B ade20k_instance_catid_mapping.txt 1KB ade20k_instance_imgCatIds.json 2.01MB prepare_coco_stuff_10k_v1.0_sem_seg.py 2KB README.md 5KB prepare_ade20k_full_sem_seg.py 52KB prepare_ade20k_pan_seg.py 13KB CONTRIBUTING.md 2KB LICENSE 19KB tools convert-pretrained-swin-model-to-d2.py 856B README.md 1KB convert-torchvision-to-d2.py 2KB MODEL_ZOO.md 25KB README.md 2KB CODE_OF_CONDUCT.md 244B .gitignore 555B mask_former test_time_augmentation.py 4KB utils misc.py 4KB __init__.py 51B config.py 3KB modeling backbone swin.py 27KB __init__.py 51B criterion.py 8KB heads per_pixel_baseline.py 9KB pixel_decoder.py 11KB mask_former_head.py 5KB __init__.py 51B __init__.py 282B transformer __init__.py 51B position_encoding.py 2KB transformer_predictor.py 6KB transformer.py 12KB matcher.py 7KB __init__.py 618B mask_former_model.py 13KB data dataset_mappers detr_panoptic_dataset_mapper.py 6KB __init__.py 51B mask_former_semantic_dataset_mapper.py 7KB mask_former_panoptic_dataset_mapper.py 6KB __init__.py 74B datasets register_mapillary_vistas.py 13KB register_ade20k_full.py 51KB __init__.py 185B register_ade20k_panoptic.py 19KB register_coco_stuff_10k.py 13KB INSTALL.md 555B资源评论
