Contamination MapVision / Model
AdaCLIP, trained on VisA + ClinicDB
- Organisation
- Huazhong University of Science and Technology et al.
- Family
- AdaCLIP · version pretrained_visa_clinicdb.pth
- Released
- 2024-07-22 (arXiv v1 submission date (https://arxiv.org/abs/2407.15795))
- Checkpoints
- weights/pretrained_visa_clinicdb.pth
- Weights
- Open · MIT
- Links
- PaperCode
Training data disclosure: Full list
Level determined for 12% of benchmark cells.
- TestTest data seen1
- SeenSeen, split unclear0
- TrainTraining data seen0
- Related data seen0
- NoNot seen0
- ?Unknown7
“The description to the utilized training set in our paper is not accurate. By default, we utilize MVTec AD & ColonDB for training, and VisA & ClinicDB are utilized for evaluations on MVTec AD & ColonDB.”
“# pre-trained from Visa and Clinicdb ckt_path="weights/pretrained_visa_clinicdb.pth"”
Built on
Everything these base models may have seen is inherited by AdaCLIP, trained on VisA + ClinicDB.
-
OpenAI CLIP
· disclosure: Partial list
Evidence
“This study employs the pre-trained CLIP (ViTL/14@336px)5 as the default backbone”
arxiv.org · paper · Section 5.1, Implementation Details · retrieved 2026-10-15 “create_model_and_transforms(backbone, image_size, pretrained='openai')”
raw.githubusercontent.com · code · method/trainer.py · retrieved 2026-10-15
Other declared training data
-
VisA · split: test · fine-tuning
Split from code: dataset classes load only the 'test' entries of meta.json.
Evidence
“The description to the utilized training set in our paper is not accurate. By default, we utilize MVTec AD & ColonDB for training, and VisA & ClinicDB are utilized for evaluations on MVTec AD & ColonDB.”
raw.githubusercontent.com · readme · Corrections · retrieved 2026-10-15 “Only utilize the test dataset for both training and testing”
raw.githubusercontent.com · code · dataset/base_dataset.py · retrieved 2026-10-15 -
ClinicDB · split: test · fine-tuning
Medical (CVC-ClinicDB) dataset.
Evidence
“The description to the utilized training set in our paper is not accurate. By default, we utilize MVTec AD & ColonDB for training, and VisA & ClinicDB are utilized for evaluations on MVTec AD & ColonDB.”
raw.githubusercontent.com · readme · Corrections · retrieved 2026-10-15 “Only utilize the test dataset for both training and testing”
raw.githubusercontent.com · code · dataset/base_dataset.py · retrieved 2026-10-15
Benchmark cells
Industrial anomaly detection
Unknown MVTec AD · unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Unknown MVTec AD 2 · unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Unknown MVTec LOCO AD · unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Test data seen VisA · in fine-tuning data: its test split was used for training
Evidence (2)
“The description to the utilized training set in our paper is not accurate. By default, we utilize MVTec AD & ColonDB for training, and VisA & ClinicDB are utilized for evaluations on MVTec AD & ColonDB.”
“Only utilize the test dataset for both training and testing”
Unknown Real-IAD · unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Unknown MPDD · unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Unknown BTAD · unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Unknown KolektorSDD2 · unknown for its base model OpenAI CLIP: training data not disclosed in enough detail
Notes
Used only for evaluation on MVTec AD and ColonDB (test.sh). See adaclip-mvtec-colondb for the paper/README discrepancy about which medical dataset is paired with which industrial dataset.
Sources
- AdaCLIP: Adapting CLIP with Hybrid Learnable Prompts for Zero-Shot Anomaly Detection (ECCV 2024) · paper · retrieved 2026-10-15
- AdaCLIP README · readme · retrieved 2026-10-15
- dataset/base_dataset.py · code · retrieved 2026-10-15
- test.sh · code · retrieved 2026-10-15
- method/trainer.py · code · retrieved 2026-10-15