KO | EN

GitLite

Minimalist Open Source Repository Search Engine

gitlite — search

#dsh #dsh-plugin #ai-agent #developer-tools #codex #typescript #cli #deepseek-harness #claude-code #ai-agents #mcp #rust

AirDet

★ 86 Open GitHub ↗

Full conference version of AirDet: Few-Shot Detection without Fine-tuning for Autonomous Exploration

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# AirDet: Few-Shot Detection without Fine-tuning for Autonomous Exploration

#### Bowen Li, Chen Wang, Pranay Reddy, Seungchan Kim, and Sebastian Scherer*

#### European Conference on Computer Vision (ECCV2022)

#### [Project web](https://jaraxxus-me.github.io/ECCV2022_AirDet/)

## Abstract

Few-shot object detection has attracted increasing attention and rapidly progressed in recent years. However, the requirement of an exhaustive offline fine-tuning stage in existing methods is time-consuming and significantly hinders their usage in online applications such as autonomous exploration of low-power robots. We find that their major limitation is that the little but valuable information from a few support images is not fully exploited. To solve this problem, we propose a brand new architecture, AirDet, and surprisingly find that, by learning *class-agnostic relation* with the support images in all modules, including cross-scale object proposal network, shots aggregation module, and localization network, AirDet without fine-tuning achieves comparable or even better results than the exhaustively fine-tuned methods, reaching up to **30-40%** improvements. We also present solid results of onboard tests on real-world exploration data from the DARPA Subterranean Challenge, which strongly validate the feasibility of AirDet in robotics. To the best of our knowledge, AirDet is the first feasible few-shot detection method for autonomous exploration of low-power robots. The source code and pre-trained models are released.

## Overview

We provide official implementation here to reproduce the results **w/o** fine-tuning of ResNet101 backbone on:
- [x] COCO-2017 validation
- [x] VOC-2012 validation dataset

## Installation

Please create a python environment including:

Python                  3.6.9

numpy                   1.19.2

detectron2              0.2 (higher version is also OK)

CUDA compiler           CUDA 10.2

PyTorch                 1.5.1

Pillow                  8.3.1

torchvision             0.6.0

fvcore                  0.1.5

cv2                     4.5.4

We also provide the official [docker image](https://hub.docker.com/repository/docker/bowenli1024/fewx) (v4) for faster reproduction.

ROS wrapper is also provided [here](https://github.com/Jaraxxus-Me/AirDet_ROS).

## Dataset Preparation

### 1. Download official datasets

[MS COCO 2017](https://cocodataset.org/#home)

[PASCAL VOC](http://host.robots.ox.ac.uk/pascal/VOC/)

[COCO format VOC annotations](https://s3.amazonaws.com/images.cocodataset.org/external/external_PASCAL_VOC.zip)

Expected dataset Structure:

```shell
coco/
  annotations/
    instances_{train,val}2017.json
    person_keypoints_{train,val}2017.json
  {train,val}2017/
```

```shell
VOC20{12}/
  annotations/
  	json files
  JPEGImages/
```

### 2. Generate supports

Download and unzip support (COCO json files) [MEGA](https://mega.nz/file/QEETwCLJ#A8m0R7NhJ-MUNuT1fhzEgRIg6t5R69u5rAaBHTsqgUw)/[BaiduNet](https://pan.baidu.com/s/1cFtwrWAwTotwZKbXYyzjEA)(pwd:1134) in

```shell
datasets/
  coco/
    new_annotations/
```

Download and unzip support (VOC json files) [MEGA](https://mega.nz/file/BBcjjYwY#1S3Utg99D_WyfzN5qq0UfeuFrlh7Eum2jZs9U7GHhJY)/[BaiduNet](https://pan.baidu.com/s/1vPZmKKue4CAZQVzOnBUs-A)(pwd:1134) in

```shell
datasets/
  voc/
    new_annotations/
```

Run the script

```shell
cd datasets
bash generate_support_data.sh
```

You may modify 4_gen_support_pool_10_shot.py line 190, 213, and 269 with different shots (default is 1 shot).

## Reproduce

### Base training

Download base [R-101 model](https://dl.fbaipublicfiles.com/detectron2/ImageNetPretrained/MSRA/R-101.pkl) in /output

start training

```shell
bash train.sh
```

We also provide official trained model [MEGA](https://mega.nz/file/1YcBWQ4R#mCtaHS9RE2pzmPlAmOAAtk-IghBNiW95oSX4Lfktw4Y)/[BaiduNet](https://pan.baidu.com/s/1iitD3hQalxpfjQcxWccV5g)(pwd:1134)

Put the model in /output/R101/

### Inference w/o fine-tuning

```shell
bash test.sh
```

You'll get the results in /log

## Citation
If our work motivates/helps your work, please cite us as:
```
@inproceedings{Li2022ECCV,
      author    = {Li, Bowen and Wang, Chen and Reddy, Pranay and Kim, Seungchan and Scherer, Sebastian},
      title     = {AirDet: Few-Shot Detection without Fine-tuning for Autonomous Exploration},
      booktitle = {Proceedings of the European Conference on Computer Vision (ECCV)},
      year      = {2022}
  }
```

## Acknowledgement

Our code is built on top of [FewX](https://github.com/fanq15/FewX), we express our sincere gratitude for the authors.

AirDet

Explore Similar Repositories

Was this content helpful?

Recent Feedback

Download README