Version: v2602

Environment Setup

AcuiRT Environment Setup

Install AIBooster and related packages by referring to the How to set up the environment outside the recommended environment.

DETR Environment Setup

Clone the aibooster-examples repository and change the working directory into DETR.

git clone -b 0.5.0 https://github.com/fixstars/aibooster-examples && cd aibooster-examples/intelligence/acuirt/detr/baseline

Install the packages required by DETR.
```
pip install -r requirements.txt
```

Dataset Preparation

Download the evaluation dataset for the COCO Dataset. Please specify an arbitrary path for /path/to/dataset.

wget http://images.cocodataset.org/annotations/annotations_trainval2017.zip
wget http://images.cocodataset.org/zips/val2017.zip
unzip annotations_trainval2017.zip -d /path/to/dataset
unzip val2017.zip -d /path/to/dataset

Please confirm that the dataset structure is as follows.

coco
├── annotations
│   ├── captions_train2017.json
│   ├── captions_val2017.json
│   ├── instances_train2017.json
│   ├── instances_val2017.json
│   ├── instances_val2017_subset.json
│   ├── person_keypoints_train2017.json
│   └── person_keypoints_val2017.json
└── val2017

val2017 contains 5000 images, so inference and evaluation take a long time. For simplicity, we will create a subset containing 50 randomly selected images.

python create_subset.py --val_json_path /path/to/dataset/coco/annotations/instances_val2017.json --output_json_path /path/to/dataset/coco/annotatinos/instances_val2017_subset.json

Download the pre-trained weights.

wget https://dl.fbaipublicfiles.com/detr/detr-r101-2c7b67e5.pth

Run inference.

python main.py --batch_size 1 --no_aux_loss --eval --backbone resnet101 --resume ./detr-r101-2c7b67e5.pth --coco_path /path/to/dataset/coco

It will be successful if a recognition accuracy log like the one below is output.

IoU metric: bbox
Average Precision  (AP) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.531
Average Precision  (AP) @[ IoU=0.50      | area=   all | maxDets=100 ] = 0.727
Average Precision  (AP) @[ IoU=0.75      | area=   all | maxDets=100 ] = 0.560
Average Precision  (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.300
Average Precision  (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.553
Average Precision  (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.720
Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=  1 ] = 0.428
Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets= 10 ] = 0.625
Average Recall     (AR) @[ IoU=0.50:0.95 | area=   all | maxDets=100 ] = 0.648
Average Recall     (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.394
Average Recall     (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.655
Average Recall     (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = 0.814

AcuiRT Environment Setup​

DETR Environment Setup​

AcuiRT Environment Setup

DETR Environment Setup