Data

Tools

Indexes

Applications

Experiments

Open Source Science

sd-dert

The official implementation of SD DETR

https://github.com/kai271828/sd-dert

Science Score: 44.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ✓
    CITATION.cff file
    Found CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ○
    DOI references
  • ○
    Academic publication links
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (11.8%) to scientific vocabulary
Last synced: 6 months ago · JSON representation ·

Repository

The official implementation of SD DETR

Basic Info
  • Host: GitHub
  • Owner: kai271828
  • License: apache-2.0
  • Language: Python
  • Default Branch: main
  • Size: 12.5 MB
Statistics
  • Stars: 1
  • Watchers: 1
  • Forks: 1
  • Open Issues: 1
  • Releases: 0
Created over 2 years ago · Last pushed over 2 years ago
Metadata Files
Readme License Citation

README.md

SD DETR

This is the official implementation of the APSIPA ASC 2023 paper, "A Transformer-Based Framework for Tiny Object Detection".

Installation

We have tested the following versions of OS and softwares:

  • OS: Ubuntu 22.04
  • GPU: Tesla V100
  • CUDA: 12.1
  • GCC(G++): 11.3.0
  • PyTorch: 2.0.0
  • TorchVision: 0.15.1
  • MMCV: 2.0.1
  • MMDetection: 3.0.0

Install

This repository is based on the MMDetection. Please refer to installation instructions of MMDetection.

shell pip install -U openmim mim install mmengine mim install "mmcv>=2.0.0"

shell git clone https://github.com/kai271828/SD-DERT.git cd SD-DERT pip install -v -e .

```shell

Install cocoapi

pip install "git+https://github.com/jwwangchn/cocoapi-aitod.git#subdirectory=aitodpycocotools"

you may need the following library

sudo apt update && sudo apt install libgl1-mesa-glx

```

Prepare datasets

Please refer to AI-TOD for AI-TOD dataset.

If your folder structure is different, you may need to change the corresponding paths in config files (configs/base/datasets/aitoddetection.py, configs/base/datasets/aitodv2detection.py).

shell home/u2339555 │ ├── AITOD │ ├── aitod │ │ ├── annotations │ │ │ │─── aitod_train.json │ │ │ │─── aitodv2_train.json │ │ │ │─── ... │ │ ├── trainval │ │ │ │─── ***.png │ │ │ │─── ***.png │ │ ├── test │ │ │ │─── ***.png │ │ │ │─── ***.png

Run

Our config files are in configs/SOD.

Please see MMDetection full tutorials Train & Test for more details.

Training on a single GPU

The basic usage is as follows. Note that the lr=0.02 in config file needs to be lr=0.02 / 8 for training on single GPU.

shell python tools/train.py configs/SOD/AITODv2_SD-DETR_2stages_NWD_60e.py

Training on multiple GPUs

The basic usage is as follows.

shell bash ./tools/dist_train.sh configs/SOD/AITODv2_SD-DETR_2stages_NWD_60e.py 8

Inference

The basic usage is as follows.

shell python tools/test.py configs/SOD/AITODv2_SD-DETR_2stages_NWD_60e.py ~/result/epoch_60.pth

Citation

BibTeX @inproceedings{SD-DETR, title={A Transformer-Based Framework for Tiny Object Detection}, author={Yi-Kai Liao, Gong-Si Lin and Mei-Chen Yeh}, booktitle={Asia Pacific Signal and Information Processing Association Annual Summit and Conference}, year={2023}, }

Owner

  • Name: Kyle Liao
  • Login: kai271828
  • Kind: user

Citation (CITATION.cff) 

cff-version: 1.2.0
message: "If you use this software, please cite it as below."
authors:
  - name: "MMDetection Contributors"
title: "OpenMMLab Detection Toolbox and Benchmark"
date-released: 2018-08-22
url: "https://github.com/open-mmlab/mmdetection"
license: Apache-2.0

GitHub Events

Total
  • Watch event: 1
Last Year
  • Watch event: 1