ixrt

YOLOF (ixRT)

Model Description

YOLOF is a lightweight object detection model that focuses on single-level feature maps for detection and enhances feature representation using dilated convolution modules. With a simple and efficient structure, it is well-suited for real-time object detection tasks.

Supported Environments

GPU	IXUCA SDK	Release	Branch
MR-V100	4.4.0	26.03	release/26.03
MR-V100	4.3.0	25.12	release/25.12

Note: 请切换到与您的 SDK 版本对应的 Release 分支进行测试。请勿直接在 master 分支上运行测试，因为 master 分支可能包含与您的本地 SDK 版本不兼容的最新更改。

切换分支命令示例：git checkout release/26.03

Model Preparation

Prepare Resources

Pretrained model: https://download.openmmlab.com/mmdetection/v2.0/yolof/yolof_r50_c5_8x8_1x_coco/yolof_r50_c5_8x8_1x_coco_20210425_024427-8e864411.pth

Dataset:

• https://github.com/ultralytics/assets/releases/download/v0.0.0/coco2017labels.zip to download the labels dataset.
• http://images.cocodataset.org/zips/val2017.zip to download the validation dataset.
• http://images.cocodataset.org/zips/train2017.zip to download the train dataset.

unzip -q -d ./ coco2017labels.zip
unzip -q -d ./coco/images/ train2017.zip
unzip -q -d ./coco/images/ val2017.zip

coco
├── annotations
│   └── instances_val2017.json
├── images
│   ├── train2017
│   └── val2017
├── labels
│   ├── train2017
│   └── val2017
├── LICENSE
├── README.txt
├── test-dev2017.txt
├── train2017.cache
├── train2017.txt
├── val2017.cache
└── val2017.txt

Install Dependencies

Contact the Iluvatar administrator to get the missing packages:

• mmcv-*.whl

pip3 install -r requirements.txt

Model Conversion

mkdir -p checkpoints/

# download the weight from the recommend link
wget https://download.openmmlab.com/mmdetection/v2.0/yolof/yolof_r50_c5_8x8_1x_coco/yolof_r50_c5_8x8_1x_coco_20210425_024427-8e864411.pth

# export onnx model
python3 export.py --weight yolof_r50_c5_8x8_1x_coco_20210425_024427-8e864411.pth --cfg ../../ixrt_common/yolof_r50-c5_8xb8-1x_coco.py --output checkpoints/yolof.onnx

Model Inference

export PROJ_DIR=./
export DATASETS_DIR=./coco/
export CHECKPOINTS_DIR=./checkpoints
export RUN_DIR=../../ixrt_common

FP16

# Accuracy
bash scripts/infer_yolof_fp16_accuracy.sh
# Performance
bash scripts/infer_yolof_fp16_performance.sh

Model Results

Model	BatchSize	Precision	FPS	IOU@0.5	IOU@0.5:0.95
YOLOF	32	FP16	331.06	0.527	0.343

References

• mmdetection