AirCode: A Robust Object Encoding Method

Overview

AirCode

This repo contains source codes for the arXiv preprint "AirCode: A Robust Object Encoding Method"

Demo

Object matching comparison when the objects are non-rigid and the view is changed, left is the result of our method while right is the result of NetVLAD

Relocalization on KITTI datasets

Dependencies

  • Python
  • PyTorch
  • OpenCV
  • Matplotlib
  • NumPy
  • Yaml

Data

Four datasets are used in our experiments.

KITTI Odometry

For relocalization experiment. Three sequences are selected, and they are "00", "05" and "06".

KITTI Tracking

For multi-object matching experiment. Four sequences are selected, and they are "0002", "0003", "0006", "0010".

VOT Datasets

For single-object matching experiment. We select three sequences from VOT2019 datasets and they are "bluecar", "bus6" and "humans_corridor_occ_2_A", because the tracked objects in these sequences are included in coco datasets, which are the data we used to train mask-rcnn.

OTB Datasets

For single-object matching experiment. We select five sequences and they are "BlurBody", "BlurCar2", "Human2", "Human7" and "Liquor".

Examples

Relocalization on KITTI Datasets

  1. Extract object descrptors

    python experiments/place_recogination/online_relocalization.py -c config/experiment_tracking.yaml -g 1 -s PATH_TO_SAVE_MIDDLE_RESULTS -d PATH_TO_DATASET -m PATH_TO_MODELS
    
  2. Compute precision-recall curves

    python experiments/place_recogination/offline_process.py -c config/experiment_tracking.yaml -g 1 -d PATH_TO_DATASET -n PATH_TO_MIDDLE_RESULTS -s PATH_TO_SAVE_RESULTS
    
  3. Compute top-K relocalization results

    python experiments/place_recogination/offline_topK.py -c config/experiment_tracking.yaml -g 1 -d PATH_TO_DATASET -n PATH_TO_MIDDLE_RESULTS -s PATH_TO_SAVE_RESULTS
    

Object Matching on OTB, VOT or KITTI Tracking Datasets

  • Run multi-object matching experiment in KITTI Tracking Datasets Modify the config file and run

    python experiments/object_tracking/object_tracking.py -c config/experiment_tracking.yaml -g 1 -s PATH_TO_SAVE_RESULTS -d PATH_TO_DATASET -m PATH_TO_MODELS 
    
  • Run single-object matching experiment in OTB or VOT Datasets Modify the config file and run

    python experiments/object_tracking/single_object_tracking.py -c config/experiment_tracking.yaml -g 1 -s PATH_TO_SAVE_RESULTS -d PATH_TO_DATASET -m PATH_TO_MODELS 
    
You might also like...
PyTorch implementation of Rethinking Positional Encoding in Language Pre-training
PyTorch implementation of Rethinking Positional Encoding in Language Pre-training

TUPE PyTorch implementation of Rethinking Positional Encoding in Language Pre-training. Quickstart Clone this repository. git clone https://github.com

Incremental Transformer Structure Enhanced Image Inpainting with Masking Positional Encoding (CVPR2022)
Incremental Transformer Structure Enhanced Image Inpainting with Masking Positional Encoding (CVPR2022)

Incremental Transformer Structure Enhanced Image Inpainting with Masking Positional Encoding by Qiaole Dong*, Chenjie Cao*, Yanwei Fu Paper and Supple

A Robust Non-IoU Alternative to Non-Maxima Suppression in Object Detection
A Robust Non-IoU Alternative to Non-Maxima Suppression in Object Detection

Confluence: A Robust Non-IoU Alternative to Non-Maxima Suppression in Object Detection 1. 介绍 用以替代 NMS,在所有 bbox 中挑选出最优的集合。 NMS 仅考虑了 bbox 的得分,然后根据 IOU 来

[ECCVW2020] Robust Long-Term Object Tracking via Improved Discriminative Model Prediction (RLT-DiMP)
[ECCVW2020] Robust Long-Term Object Tracking via Improved Discriminative Model Prediction (RLT-DiMP)

Feel free to visit my homepage Robust Long-Term Object Tracking via Improved Discriminative Model Prediction (RLT-DIMP) [ECCVW2020 paper] Presentation

 Robust Instance Segmentation through Reasoning about Multi-Object Occlusion [CVPR 2021]
Robust Instance Segmentation through Reasoning about Multi-Object Occlusion [CVPR 2021]

Robust Instance Segmentation through Reasoning about Multi-Object Occlusion [CVPR 2021] Abstract Analyzing complex scenes with DNN is a challenging ta

Code release for our paper,
Code release for our paper, "SimNet: Enabling Robust Unknown Object Manipulation from Pure Synthetic Data via Stereo"

SimNet: Enabling Robust Unknown Object Manipulation from Pure Synthetic Data via Stereo Thomas Kollar, Michael Laskey, Kevin Stone, Brijen Thananjeyan

object detection; robust detection; ACM MM21 grand challenge; Security AI Challenger Phase VII
object detection; robust detection; ACM MM21 grand challenge; Security AI Challenger Phase VII

赛题背景 在商品知识产权领域,知识产权体现为在线商品的设计和品牌。不幸的是,在每一天,存在着非法商户通过一些对抗手段干扰商标识别来逃避侵权,这带来了很高的知识产权风险和财务损失。为了促进先进的多媒体人工智能技术的发展,以保护企业来之不易的创作和想法免受恶意使用和剽窃,因此提出了鲁棒性标识检测挑战赛

Code and models for ICCV2021 paper
Code and models for ICCV2021 paper "Robust Object Detection via Instance-Level Temporal Cycle Confusion".

Robust Object Detection via Instance-Level Temporal Cycle Confusion This repo contains the implementation of the ICCV 2021 paper, Robust Object Detect

Official code for 'Robust Siamese Object Tracking for Unmanned Aerial Manipulator' and offical introduction to UAMT100 benchmark
Official code for 'Robust Siamese Object Tracking for Unmanned Aerial Manipulator' and offical introduction to UAMT100 benchmark

SiamSA: Robust Siamese Object Tracking for Unmanned Aerial Manipulator Demo video 📹 Our video on Youtube and bilibili demonstrates the evaluation of

Comments
  • how can I get *.pth files?

    how can I get *.pth files?

    Hello, I am a beginner. When I run python experiments/place_recogination/online_relocalization.py -c config/experiment_tracking.yaml -g 1 -s results/ -d /media/jixingwu/datasetj/KITTI/Odom/data_odometry_color/sequences -m models/, points_model.pth file is needed. So how can I get it? Thank you!

    opened by jixingwu 5
  • Unable to load model under CPU-only configuration

    Unable to load model under CPU-only configuration

    Hi, I want to run object tracking on KITTI tracking datasets with only CPU using the following terminal prompt:

      python experiments/object_tracking/object_tracking.py -c config/experiment_tracking.yaml -g 1 -s ./results -d /data/datasets/SLAM_dataset/training/ -m ./weights
    

    with configuration in object_tracking.py updated with

    configs['use_gpu'] = 0
    

    However, when running with the configuration above with gcn_model.pth, maskrcnn_model.pth, points_model.pth model files in release v2.0.0, the following error occurs:

    (aircode) yutianc@bender:~/workspace/AirCode$ python experiments/object_tracking/object_tracking.py -c config/experiment_tracking.yaml -g 1 -s ./results -d /data/datasets/SLAM_dataset/training/ -m ./weights
    experiments/object_tracking/object_tracking.py:371: YAMLLoadWarning: calling yaml.load() without Loader=... is deprecated, as the default Loader is unsafe. Please read https://msg.pyyaml.org/load for full details.
      configs = yaml.load(configs)
    Traceback (most recent call last):
      File "experiments/object_tracking/object_tracking.py", line 384, in <module>
        main()
      File "experiments/object_tracking/object_tracking.py", line 381, in main
        show_object_tracking(configs)
      File "experiments/object_tracking/object_tracking.py", line 272, in show_object_tracking
        superpoint_model = build_superpoint_model(configs, requires_grad=False)
      File "./model/build_model.py", line 101, in build_superpoint_model
        model.load_state_dict(model_dict)
      File "/home/yutianc/minicondas/envs/aircode/lib/python3.7/site-packages/torch/nn/modules/module.py", line 1052, in load_state_dict
        self.__class__.__name__, "\n\t".join(error_msgs)))
    RuntimeError: Error(s) in loading state_dict for VggLike:
            Unexpected key(s) in state_dict: "module.pretrained_net.features.0.weight", "module.pretrained_net.features.0.bias", "module.pretrained_net.features.2.weight", "module.pretrained_net.features.2.bias", "module.pretrained_net.features.5.weight", "module.pretrained_net.features.5.bias", "module.pretrained_net.features.7.weight", "module.pretrained_net.features.7.bias", "module.pretrained_net.features.10.weight", "module.pretrained_net.features.10.bias", "module.pretrained_net.features.12.weight", "module.pretrained_net.features.12.bias", "module.pretrained_net.features.14.weight", "module.pretrained_net.features.14.bias", "module.pretrained_net.features.17.weight", "module.pretrained_net.features.17.bias", "module.pretrained_net.features.19.weight", "module.pretrained_net.features.19.bias", "module.pretrained_net.features.21.weight", "module.pretrained_net.features.21.bias", "module.pretrained_net.features.24.weight", "module.pretrained_net.features.24.bias", "module.pretrained_net.features.26.weight", "module.pretrained_net.features.26.bias", "module.pretrained_net.features.28.weight", "module.pretrained_net.features.28.bias", "module.convPa.weight", "module.convPa.bias", "module.bnPa.weight", "module.bnPa.bias", "module.bnPa.running_mean", "module.bnPa.running_var", "module.bnPa.num_batches_tracked", "module.convPb.weight", "module.convPb.bias", "module.bnPb.weight", "module.bnPb.bias", "module.bnPb.running_mean", "module.bnPb.running_var", "module.bnPb.num_batches_tracked", "module.convDa.weight", "module.convDa.bias", "module.bnDa.weight", "module.bnDa.bias", "module.bnDa.running_mean", "module.bnDa.running_var", "module.bnDa.num_batches_tracked", "module.convDb.weight", "module.convDb.bias", "module.bnDb.weight", "module.bnDb.bias", "module.bnDb.running_mean", "module.bnDb.running_var", "module.bnDb.num_batches_tracked".
    

    Running object_tracking.py with CUDA seems to load models successfully. Is there something wrong with the model loading when GPU is disabled?

    opened by MarkChenYutian 4
  • Why RGB image is converted into grayscale image with 3 channels?

    Why RGB image is converted into grayscale image with 3 channels?

    Hi, I'm trying to use AirCode to do object matching on complete KITTI sequences and I'm reading the code in experiments/show_object_matching.py.

    While reading the code, I noticed that the current code is reading RGB image sequence, convert it into grayscale image, and then duplicate the image into 3-channel each with same value (as following):

    https://github.com/wang-chen/AirCode/blob/5e23e9f5322d2e4ee119d5326a6b6112cef0e6bd/experiments/show_object_matching/show_object_matching.py#L172-L176

    I'm a bit unsure about the reason why this operation is performed here as the original RGB image should contain more information about the object comparing to grayscale image. For instance, it should be easier to distinguish objects with different color but similar shape if the RGB value is preserved.

    opened by MarkChenYutian 2
Owner
Chen Wang
I am engaged in delivering simple and efficient source code.
Chen Wang
git《FSCE: Few-Shot Object Detection via Contrastive Proposal Encoding》(CVPR 2021) GitHub: [fig8]

FSCE: Few-Shot Object Detection via Contrastive Proposal Encoding (CVPR 2021) This repo contains the implementation of our state-of-the-art fewshot ob

null 233 Dec 29, 2022
LBK 35 Dec 26, 2022
Tools to create pixel-wise object masks, bounding box labels (2D and 3D) and 3D object model (PLY triangle mesh) for object sequences filmed with an RGB-D camera.

Tools to create pixel-wise object masks, bounding box labels (2D and 3D) and 3D object model (PLY triangle mesh) for object sequences filmed with an RGB-D camera. This project prepares training and testing data for various deep learning projects such as 6D object pose estimation projects singleshotpose, as well as object detection and instance segmentation projects.

null 305 Dec 16, 2022
Multi-Scale Vision Longformer: A New Vision Transformer for High-Resolution Image Encoding

Vision Longformer This project provides the source code for the vision longformer paper. Multi-Scale Vision Longformer: A New Vision Transformer for H

Microsoft 209 Dec 30, 2022
Distance Encoding for GNN Design

Distance-encoding for GNN design This repository is the official PyTorch implementation of the DEGNN and DEAGNN framework reported in the paper: Dista

null 172 Nov 8, 2022
Relative Positional Encoding for Transformers with Linear Complexity

Stochastic Positional Encoding (SPE) This is the source code repository for the ICML 2021 paper Relative Positional Encoding for Transformers with Lin

Antoine Liutkus 48 Nov 16, 2022
[ACMMM 2021 Oral] Enhanced Invertible Encoding for Learned Image Compression

InvCompress Official Pytorch Implementation for "Enhanced Invertible Encoding for Learned Image Compression", ACMMM 2021 (Oral) Figure: Our framework

null 96 Nov 30, 2022
Pytorch implementation of CVPR2020 paper “VectorNet: Encoding HD Maps and Agent Dynamics from Vectorized Representation”

VectorNet Re-implementation This is the unofficial pytorch implementation of CVPR2020 paper "VectorNet: Encoding HD Maps and Agent Dynamics from Vecto

null 120 Jan 6, 2023
Eth brownie struct encoding example

eth-brownie struct encoding example Overview This repository contains an example of encoding a struct, so that it can be used in a function call, usin

Ittai Svidler 2 Mar 4, 2022