Development of Industrial IoT System for Anomaly Detection in Smart Factory


CUPSNBOTTLES is an object data set, recorded by a mobile service robot. There are 10 object classes, each with a varying number of samples. Additionally, there is a clutter class, containing samples where the object detector failed.


Download and extract the ZIP file containing all files. There is python code available (under 'scripts') to easily load the data set. Other programming languages should also handle .jpg, .hdf and .csv files for easy access. For easy access with python, a pickle dump file has been added. This has no extra information compared to the .csv file.


This dataset was created for the following paper:


PX4 Autopilot (v1.10.1 stable) ( is used for all experiments, running on Pixhawk 4 flight controller for HITL. QGroundControl (v4.0.9) is used for GCS (

Telemetry data is contained in TLOG files (

Full flight data is contained in ULOG files (

It is useful to use ulog2csv to extract more information in CSV format:

GPS spoofing attacks are carried out for 30 seconds. The attacks are done by stopping normal GPS communications, then injecting false readings via Gazebo. This is done by a modification of sitl_gazebo in gazebo_gps_plugin.cpp (


Along with the increasing use of unmanned aerial vehicles (UAVs), large volumes of aerial videos have been produced. It is unrealistic for humans to screen such big data and understand their contents. Hence methodological research on the automatic understanding of UAV videos is of paramount importance.


=================  Authors  ===========================

Lichao Mou,

Yuansheng Hua,

Pu Jin,

Xiao Xiang Zhu,


=================  Citation  ===========================

If you use this dataset for your work, please use the following citation:


  title= {{ERA: A dataset and deep learning benchmark for event recognition in aerial videos}},

  author= {Mou, L. and Hua, Y. and Jin, P. and Zhu, X. X.},

  journal= {IEEE Geoscience and Remote Sensing Magazine},

  year= {in press}



==================  Notice!  ===========================

This dataset is ONLY released for academic uses. Please do not further distribute the dataset on other public websites.


This is a dataset having paired thermal-visual images collected over 1.5 years from different locations in Chitrakoot, India and Prayagraj, India. The images can be broadly classified into greenery, urban, historical buildings and crowd data.

The crowd data was collected from the Maha Kumbh Mela 2019, Prayagraj, which is the largest religious fair in the world and is held every 6 years.



The images are classified according to the thermal imager they were used to capture them with.

The SONEL thermal images are inside register_sonel.

The FLIR images are in register_flir and register_flir_old. There are 2 image zip files because FLIR thermal imagers reuse the image names after a certain limit.

The unregistered images are kept as files inside each base zip as unreg folders.


The work associated with this database, which details the registration method, the overall logic behind the creation of this database, resizing factors and the reason why there are unregistered images, is a work on thermal image colorization has been submited to IEEE for consideration, and is currently pre printed and available on arXiv.

We ask that you refer to this work when using this database for your work.

A Novel Registration & Colorization Technique for Thermal to Cross Domain Colorized Images 


If you find any problem with the data in this dataset (missing images, wrong names, superfluous python files etc), please let us know and we will try to correct the same.


The naming classification is as follows:

·         FLIR

o   Registered images are named as <name>.jpg and <name>_color.png with the png file being the optical registered image

o   The raw files are named as FLIR<#number>.jpg and FLIR<#number+1>.jpg where the initial file is the thermal image

o   The unreg_flir folder contains just the raw files

·         SONEL

o   Registered images are named as <name>.jpg and <name>_color.png with the png file being the optical registered image

o   The raw files are named as IRI_<name>.jpg and VIS_< name >.jpg where the IRI file is the thermal image and VIS is the visual image

o   The unreg folder contains just the raw files


A Traffic Light Controller PETRI_NET (Finite State Machine) Implementation.


An implementation of FSM approach can be followed in systems whose tasks constitute a well-structured list so all states can be easily enumerated. A Traffic light controller represents a relatively complex control function


This file would need to be unzipped for access




7.8 磅




Dataset of GPS, inertial and WiFi data collected during road vehicle trips in the district of Porto, Portugal. It contains 40 trip datasets collected with a smartphone fixed on the windshield or dashboard, inside the road vehicle. The dataset was collected and used in order to develop a proof-of-concept for "MagLand: Magnetic Landmarks for Road Vehicle Localization", an approach that leverages magnetic anomalies created by existing road infrastructure as landmarks, in order to support current vehicle localization system (e.g. GNSS, dead reckoning).


Dataset is organized in folders by date.Inside each folder, it is separated in folders by collection app or equipment.Inside collection app/equipemnt folders, it is separated by sensor.For each sensor there is a time series per trip.For details about the trips, including vehicles, smartphones, apps, and dates for data collection please read "README.txt".


Collision detection (CD) is a key capability of carrier sense multiple access (CSMA) based medium access control (MAC) protocol. Applying CD, the transmitter can abort transmission immediately so that the power can be saved. This technique does not need the peer receiver to give feedback on whether there is a packet collision, and hence, the overall overhead is significantly low. The challenge, however, is to operate in transmit time and instantly detect the week colliding signal in the presence of strong self-interference (SI).


Instant collision detection (CD) can be achieved at the transmitter side more efficiently. To detect the collision, though, the device has to overcome the strong self-interference (SI) in such a way that it can listen to the channel in transmit time. This capability is feasible by in-band full-duplex (IBFD) technology, which allows two nodes to communicate concurrently over the same frequency channel. Recent works have shown the network-level benefits of using IBFD for collision detection, in the sense of power efficiency, throughput, and delay performance. By any means, the performance of these MAC protocols highly depends on the rapidity and precision of the CD method, although the collision detection in this context has still not been investigated thoroughly. By leveraging multiple hidden convolutional layers, modern machine learning techniques have confirmed their effectiveness in a wide range of applications, such as automatic image recognition, and network optimization. Motivated by its remarkable success in various fields as well as its real-time functionality, in this work we investigate whether a convolutional neural network (CNN) can be exploited to accelerate CD without sacrificing the detection accuracy. Meanwhile, we realize that the CD problem can be mapped to traditional SNR estimation problem. When there is a collision, the signal SNR will drop. Lots of domain knowledge are there with regard to signal demodulation and SNR estimation. On the contrary, CNN could be regarded as a kind of domain-specific knowledge less method. It will be interesting to see the performance comparison between the two methodologies. This kind of comparison will inspire the research community to study further about how should we combine the domain-specific knowledge (DSK) with CNN. Besides, to encourage future studies, we offer free access to the dataset and programs in IEEE DataPort, which allows researchers to reproduce our results out of the box or investigate different approaches.


This dataset features cooking activities with recipes and gestures labeled. The data has been collected using two smartphones (right arm and left hip), two smartwatches (both wrists) and one motion capture system with 29 markers. There were 4 subjects who prepared 3 recipes (sandwich, fruit salad, cereal) 5 times each. The subjects followed a script for each recipe but acted as naturally as possible


You can use our tutorials to get started