FoundAD
[ICLR 2026] The implementation of the paper Foundation Visual Encoders Are Secretly Few-Shot Anomaly Detectors
Language: python
Author: GitHub Repos (@github-repos)
0 stars · 0 views
Files
- vitl14_reg4_pretrain.yaml (yaml)
- foundad (txt)
- fsdp (txt)
- run (txt)
- train (txt)
- eval (txt)
- eval (txt)
- decode_heads (txt)
- depth (txt)
- depther (txt)
- losses (txt)
- backbones (txt)
- decode_heads (txt)
- core (txt)
- utils (txt)
- samplers (txt)
- ops (txt)
- modules (txt)
- distributed (txt)
- loss (txt)
- hub (txt)
- depth (txt)
- datasets (txt)
- vision_backbone (txt)
- scripts (txt)
- additional-datasets (txt)
- prompting (txt)
- datasets (txt)
- util (txt)
- conf (txt)
- training (txt)
- assets (txt)
- mvtec (txt)
- bottle (txt)
- visa (txt)
- chewinggum (txt)
- ok (txt)
- ko (txt)
- README.md (md)
- src (txt)
- utils (txt)
- main.py (py)
- synthesis.py (py)
- logging.py (py)
- dinov2 (txt)
- models (txt)
- AD.py (py)
- metrics.py (py)
- utils (txt)
- vision_transformer.py (py)
- __init__.py (py)
- config.py (py)
- param_groups.py (py)
- dtype.py (py)
- logging (txt)
- cluster.py (py)
- train (txt)
- __init__.py (py)
- helpers.py (py)
- ssl_meta_arch.py (py)
- __init__.py (py)
- __init__.py (py)
- train.py (py)
- __init__.py (py)
- submit.py (py)
- __init__.py (py)
- train.py (py)
- knn.py (py)
- linear.py (py)
- linear.py (py)
- log_regression.py (py)
- __init__.py (py)
- segmentation (txt)
- models (txt)
- backbones (txt)
- __init__.py (py)
- __init__.py (py)
- vision_transformer.py (py)
- linear_head.py (py)
- utils (txt)
- __init__.py (py)
- colormaps.py (py)
- __init__.py (py)
- hooks (txt)
- __init__.py (py)
- knn.py (py)
- optimizer.py (py)
- models (txt)
- losses (txt)
- builder.py (py)
- __init__.py (py)
- gradientloss.py (py)
- backbones (txt)
- encoder_decoder.py (py)
- __init__.py (py)
- vision_transformer.py (py)
- decode_heads (txt)
- base.py (py)
- __init__.py (py)
- decode_head.py (py)
- dpt_head.py (py)
- __init__.py (py)
- linear_head.py (py)
- ops (txt)
- segmentation_m2f (txt)
- __init__.py (py)
- wrappers.py (py)
- log_regression.py (py)
- __init__.py (py)
- models (txt)
- utils (txt)
- builder.py (py)
- assigner.py (py)
- __init__.py (py)
- transformer.py (py)
- point_sample.py (py)
- segmentors (txt)
- __init__.py (py)
- __init__.py (py)
- match_costs.py (py)
- __init__.py (py)
- dice_loss.py (py)
- encoder_decoder_mask2former.py (py)
- cross_entropy_loss.py (py)
- vit_adapter.py (py)
- drop_path.py (py)
- plugins (txt)
- adapter_modules.py (py)
- __init__.py (py)
- __init__.py (py)
- __init__.py (py)
- vit.py (py)
- msdeformattn_pixel_decoder.py (py)
- __init__.py (py)
- __init__.py (py)
- box (txt)
- __init__.py (py)
- misc.py (py)
- dist_utils.py (py)
- sampling_result.py (py)
- builder.py (py)
- mask_pseudo_sampler.py (py)
- __init__.py (py)
- base_sampler.py (py)
- anchor (txt)
- mask_sampling_result.py (py)
- point_generator.py (py)
- builder.py (py)
- __init__.py (py)
- __init__.py (py)
- __init__.py (py)
- ms_deform_attn.py (py)
- metrics.py (py)
- layers (txt)
- rope_position_encoding.py (py)
- setup.py (py)
- utils.py (py)
- __init__.py (py)
- patch_embed.py (py)
- block.py (py)
- dino_head.py (py)
- mlp.py (py)
- layer_scale.py (py)
- swiglu_ffn.py (py)
- attention.py (py)
- __init__.py (py)
- ibot_patch_loss.py (py)
- dino_clstoken_loss.py (py)
- __init__.py (py)
- __init__.py (py)
- koleo_loss.py (py)
- backbones.py (py)
- encoder_decoder.py (py)
- decode_heads.py (py)
- ops.py (py)
- __init__.py (py)
- classifiers.py (py)
- configs (txt)
- utils.py (py)
- train (txt)
- vitl16_short.yaml (yaml)
- vitl14.yaml (yaml)
- eval (txt)
- __init__.py (py)
- vitg14_reg4_pretrain.yaml (yaml)
- vitg14.yaml (yaml)
- vitg14_pretrain.yaml (yaml)
- vitb14_pretrain.yaml (yaml)
- vits14_reg4_pretrain.yaml (yaml)
- vitb14_reg4_pretrain.yaml (yaml)
- vitl14_pretrain.yaml (yaml)
- vits14_pretrain.yaml (yaml)
- ssl_default_config.yaml (yaml)
- dataset.py (py)
- vit_inference.py (py)
- Makefile (txt)
- preprocess.py (py)
- generate.py (py)
- prismatic (txt)
- models (txt)
- lrv_instruct.py (py)
- pretrain.py (py)
- lvis_instruct_4v.py (py)
- __init__.py (py)
- materialize.py (py)
- backbones (txt)
- llm (txt)
- llama2.py (py)
- __init__.py (py)
- mistral.py (py)
- phi.py (py)
- __init__.py (py)
- phi_prompter.py (py)
- llama2_chat_prompter.py (py)
- mistral_instruct_prompter.py (py)
- vicuna_v15_prompter.py (py)
- base_prompter.py (py)
- vision (txt)
- base_llm.py (py)
- siglip_vit.py (py)
- vlms (txt)
- prismatic.py (py)
- dinosiglip_vit.py (py)
- base_vision.py (py)
- dinov2_vit.py (py)
- dinoclip_vit.py (py)
- in1k_vit.py (py)
- registry.py (py)
- base_vlm.py (py)
- vit.py (py)
- __init__.py (py)
- overwatch (txt)
- preprocessing (txt)
- download.py (py)
- __init__.py (py)
- overwatch.py (py)
- __init__.py (py)
- datasets.py (py)
- __init__.py (py)
- torch_utils.py (py)
- data_utils.py (py)
- materialize.py (py)
- __init__.py (py)
- batching_utils.py (py)
- models.py (py)
- datasets.py (py)
- nn_utils.py (py)
- __init__.py (py)
- __init__.py (py)
- strategies (txt)
- metrics.py (py)
- ddp.py (py)
- fsdp.py (py)
- base_strategy.py (py)
- materialize.py (py)
- __init__.py (py)
- train.py (py)
- foundad.py (py)
- sample.py (py)
- configs (txt)
- app (txt)
- helper.py (py)
- test.yaml (yaml)
- train_dinov3.yaml (yaml)
- sample_few_shot.yaml (yaml)
- config.yaml (yaml)
- 000.JPG (image)
- broken_small (txt)
- broken_large (txt)
- contamination (txt)
- good (txt)
- 004.png (image)
- 000.png (image)
- icon.png (image)
- 002.png (image)
- 000.png (image)
- 006.JPG (image)
- requirements.txt (txt)
- ATTRIBUTION.md (markdown)
- load.py (py)
- setup.py (py)
- __init__.py (py)
- __init__.py (py)
- positional_encoding.py (py)
- tensors.py (py)
- clip_vit.py (py)
- __init__.py (py)
- utils.py (py)
- __init__.py (py)
- drop_path.py (py)
- __init__.py (py)
- mask2former_head.py (py)
- depthers.py (py)
- sigloss.py (py)
