This repository contains the BoTox dataset, a German dataset for analyzing offensive language and conversations in the context of criminal relevance under the German Criminal Code (StGB). It was created as part of the BoTox research project (https://botox.h-da.de/) in 2025, which aimed to investigate hate speech in a criminal law context and to detect hate speech bots.Contentpublic dataset with comment texts as csvannotation guidelines (original German version)About the DatasetThis is a German dataset on offensive language, annotated according to three predefined classes, each with several paragraphs from German criminal law. In total, the dataset contains 1,190 annotated comments from various sources. The data was compiled in 2025 and annotated by three teams of annotators after intensive training by the Frankfurt am Main Public Prosecutor's Office. The dataset contains annotations according to inter-annotator agreement. Further details can be found in the accompanying paper.Description of the columns Column NameDescriptionindexrolling indextextcomment textclass_1binary annotationclass_2binary annotationclass_3binary annotationclass_0binary annotationoutputtextual annotationCitationIf you use the dataset, please cite our respective paper A Novel Dataset for Classifying German Hate Speech Comments with Criminal Relevance", which was presented on the on the 9th Workshop on Online Abuse and Harms (WOAH) at August 1, 2025 as part of the ACL conference.ACL-Style:Vincent Kums, Florian Meyer, Luisa Pivit, Uliana Vedenina, Jonas Wortmann, Melanie Siegel and Dirk LabuddeA Novel Dataset for Classifying German Hate Speech Comments with Criminal Relevance. In Proceedings of the 9th Workshop on Online Abuse and Harms (WOAH), pages 41-52, Vienna, Austria. Association for Computational Linguistics.BibTeX:@inproceedings{kums-etal-2025-novel, title = "A Novel Dataset for Classifying {G}erman Hate Speech Comments with Criminal Relevance", author = "Kums, Vincent and Meyer, Florian and Pivit, Luisa and Vedenina, Uliana and Wortmann, Jonas and Siegel, Melanie and Labudde, Dirk", editor = "Calabrese, Agostina and de Kock, Christine and Nozza, Debora and Plaza-del-Arco, Flor Miriam and Talat, Zeerak and Vargas, Francielle", booktitle = "Proceedings of the The 9th Workshop on Online Abuse and Harms (WOAH)", month = aug, year = "2025", address = "Vienna, Austria", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2025.woah-1.4/", pages = "41--52", ISBN = "979-8-89176-105-6"}

This repository contains the BoTox dataset, a German dataset for analyzing offensive language and conversations in the context of criminal relevance under the German Criminal Code (StGB). It was created as part of the BoTox research project (https://botox.h-da.de/) in 2025, which aimed to investigate hate speech in a criminal law context and to detect hate speech bots.Contentpublic dataset with comment texts as csvannotation guidelines (original German version)About the DatasetThis is a German dataset on offensive language, annotated according to three predefined classes, each with several paragraphs from German criminal law. In total, the dataset contains 1,190 annotated comments from various sources. The data was compiled in 2025 and annotated by three teams of annotators after intensive training by the Frankfurt am Main Public Prosecutor's Office. The dataset contains annotations according to inter-annotator agreement. Further details can be found in the accompanying paper.Description of the columns Column NameDescriptionindexrolling indextextcomment textclass_1binary annotationclass_2binary annotationclass_3binary annotationclass_0binary annotationoutputtextual annotationCitationIf you use the dataset, please cite our respective paper A Novel Dataset for Classifying German Hate Speech Comments with Criminal Relevance", which was presented on the on the 9th Workshop on Online Abuse and Harms (WOAH) at August 1, 2025 as part of the ACL conference.ACL-Style:Vincent Kums, Florian Meyer, Luisa Pivit, Uliana Vedenina, Jonas Wortmann, Melanie Siegel and Dirk LabuddeA Novel Dataset for Classifying German Hate Speech Comments with Criminal Relevance. In Proceedings of the 9th Workshop on Online Abuse and Harms (WOAH), pages 41-52, Vienna, Austria. Association for Computational Linguistics.BibTeX:@inproceedings{kums-etal-2025-novel, title = "A Novel Dataset for Classifying {G}erman Hate Speech Comments with Criminal Relevance", author = "Kums, Vincent and Meyer, Florian and Pivit, Luisa and Vedenina, Uliana and Wortmann, Jonas and Siegel, Melanie and Labudde, Dirk", editor = "Calabrese, Agostina and de Kock, Christine and Nozza, Debora and Plaza-del-Arco, Flor Miriam and Talat, Zeerak and Vargas, Francielle", booktitle = "Proceedings of the The 9th Workshop on Online Abuse and Harms (WOAH)", month = aug, year = "2025", address = "Vienna, Austria", publisher = "Association for Computational Linguistics", url = "https://aclanthology.org/2025.woah-1.4/", pages = "41--52", ISBN = "979-8-89176-105-6"}

The dataset presented is a valuable resource designed for benchmarking anomaly detection algorithms in the context of the European electricity grid. It is derived from the operation of the PICASSO platform, which is the world's first system for cross-border control of secondary control reserves, aimed at enhancing efficiency in the European electricity market.Dataset OverviewData Source: PICASSO platformDuration: 202 days of operational dataControl Areas: Two control areas connected to the PICASSO platformSampling Rate: Four secondsTotal Features: 13 featuresTest Segments: 20 labeled two-hour test segments with known anomaliesKey FeaturesFeatureDescriptionaFRR activationCurrently activated Automatic Frequency Restoration ReserveCorrected DemandDemand adjusted for corrections calculated by PICASSOaFRR requestTarget aFRR activation after optimization by PICASSOControlBandPos/NegAvailable aFRR for the control areaLFCInputCurrent imbalance of the control areaFRCEACE correctionDemandCurrent imbalance including activated aFRR and mFRRCorrectionCorrection value calculated by PICASSOParticipationCMO/INParticipation status in imbalance netting / merit orderBand Limited Corrected DemandTSO demand capped by maximum available aFRRControl AreaIdentifier of the control areaAnomaly TypesThe dataset includes various anomaly types, characterized by significant deviations in the control signals. This includes discrepancies between requested and activated reserves, delays in signal responses, and other irregular behaviors identified through expert validation.SignificanceThis dataset serves to advance research in anomaly detection, providing a benchmark for developing novel algorithms to enhance grid stability. It highlights the complexities of modern energy systems and the importance of automated anomaly detection for operational safety and efficiency.The dataset is accessible for researchers aiming to improve anomaly detection methodologies within the energy sector, fostering innovation and collaboration across the industry. An detailed describtion can be found here.

The dataset presented is a valuable resource designed for benchmarking anomaly detection algorithms in the context of the European electricity grid. It is derived from the operation of the PICASSO platform, which is the world's first system for cross-border control of secondary control reserves, aimed at enhancing efficiency in the European electricity market.Dataset OverviewData Source: PICASSO platformDuration: 202 days of operational dataControl Areas: Two control areas connected to the PICASSO platformSampling Rate: Four secondsTotal Features: 13 featuresTest Segments: 20 labeled two-hour test segments with known anomaliesKey FeaturesFeatureDescriptionaFRR activationCurrently activated Automatic Frequency Restoration ReserveCorrected DemandDemand adjusted for corrections calculated by PICASSOaFRR requestTarget aFRR activation after optimization by PICASSOControlBandPos/NegAvailable aFRR for the control areaLFCInputCurrent imbalance of the control areaFRCEACE correctionDemandCurrent imbalance including activated aFRR and mFRRCorrectionCorrection value calculated by PICASSOParticipationCMO/INParticipation status in imbalance netting / merit orderBand Limited Corrected DemandTSO demand capped by maximum available aFRRControl AreaIdentifier of the control areaAnomaly TypesThe dataset includes various anomaly types, characterized by significant deviations in the control signals. This includes discrepancies between requested and activated reserves, delays in signal responses, and other irregular behaviors identified through expert validation.SignificanceThis dataset serves to advance research in anomaly detection, providing a benchmark for developing novel algorithms to enhance grid stability. It highlights the complexities of modern energy systems and the importance of automated anomaly detection for operational safety and efficiency.The dataset is accessible for researchers aiming to improve anomaly detection methodologies within the energy sector, fostering innovation and collaboration across the industry. An detailed describtion can be found here.

Dataset containing EEG and eye-tracking data of 113 participants, recorded with consumer-grade hardware.

Dataset containing EEG and eye-tracking data of 113 participants, recorded with consumer-grade hardware.

Dataset containing EEG and eye-tracking data of 113 participants, recorded with consumer-grade hardware.

The CarDS dataset is a multi-protocol in-vehicle network dataset primarily targeted at the development of Intrusion Detection Systems (IDSs). It presents both benign traffic as well as advanced attacks launched against the in-vehicle network of a modern commercial electric vehicle from 2020 consisting of 10 internal CAN buses (i.e., domains) and 6 Automotive Ethernet buses. Specifically, the dataset covers 9h07m09s of real in-vehicle network data and features 397,383,125 CAN (CAN CC and CAN FD) and 180,604,377 Automotive Ethernet messages distributed over different scenarios in 258 traces.

Abstract Developing robust drone detection systems is often constrained by the limited availability of large-scale annotated training data and the high costs associated with real-world data collection. However, synthetic data presents a promising and cost-effective solution to overcome this issue. Therefore, we present SynDroneVision, a synthetic dataset specifically designed for RGB-based drone detection in surveillance applications. Featuring diverse backgrounds, lighting conditions, and drone models, SynDroneVision offers a comprehensive training foundation for deep learning algorithms. To evaluate the dataset's effectiveness, we perform a comparative analysis across a selection of recent YOLO detection models. Our findings demonstrated that SynDroneVision is a valuable resource for real-world data enrichment, achieving notable enhancements in model performance and robustness, while significantly reducing the time and costs of real-world data acquisition.  PaperPublished in the Proceedings of the 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV2025)!SynDroneVision is presented in the paper SynDroneVision: A Synthetic Dataset for Image-Based Drone Detection by Tamara R. Lenhard, Andreas Weinmann, Kai Franke, and Tobias Koch. This work is published in the Proceedings of the 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV2025).The preprint is currently available on ArXiv: hereThe final version is now published in the proceedings of WACV 2025: here Dataset DetailsSynDroneVision comprises a total of 140,038 annotaed RGB images (131,238 for training, 8,800 for validation, and 4,000 for testing), featuring a resolution of 2560x1489 pixels. All images are recorded in a sequential manner using Unreal Engine 5.0 in combination with Colosseum. Apart from drone images, SynDroneVision also includes ~7% of background images (i.e., imag frames without drone instances).Annotation Format: Annotations (bounding boxes) are provided via text files according to the YOLO standard format Here, and represent the normalized coordinates of the bounding box center, while and denote the normalized bounding box wisth and height. In SynDroneVision, is always 0, indicating the drone class.DownloadThe SynDroneVision dataset offers around 900 GB of data dedicated to image-based drone detection. To facilitate the download process, we have partitioned the dataset into smaller sections. Specifically, we have divided the training data into 10 segments, organized by sequences.Annotations are available below, with image data accessible via the following links:Dataset SplitSequencesFile NameLinkSize (GB)Training SetSeq. 001 - 009images_train_seq001-009.zipTraining images PART 157 Seq. 010 - 018images_train_seq010-018.zipTrainng images PART 295.4 Seq. 019 - 027images_train_seq019-027.zipTraining images PART 396.2 Seq. 028 - 035images_train_seq028-035.zipTraining images PART 483.9 Seq. 036 - 043images_train_seq036-043.zipTraining images PART 577.1 Seq. 044 - 050images_train_seq044-050.zipTraining images PART 684.7 Seq. 051 - 056images_train_seq051-056.zipTraining images PART 786.8 Seq. 057 - 065images_train_seq057-065.zipTraining images PART 886.2 Seq. 066 - 070images_train_seq066-070.zipTraining images PART 975.7 Seq. 071 - 073images_train_seq071-073.zipTraining images PART 1038.5Validation SetSeq. 001 - 073images_val.zipValidation images55.2Test SetSeq. 001 - 073images_test.zipTest images26.5CitationIf you find SynDroneVision helpful in your research, we kindly ask that you cite the associated paper. Below is the citation in BibTeX format for your convenience:BibTeX:@INPROCEEDINGS{10943801, author={Lenhard, Tamara R. and Weinmann, Andreas and Franke, Kai and Koch, Tobias}, booktitle={2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)}, title={SynDroneVision: A Synthetic Dataset for Image-Based Drone Detection}, year={2025}, volume={}, number={}, pages={7637-7647}, doi={10.1109/WACV61041.2025.00742}}SynDroneVision uses Unreal® Engine. Unreal® is a trademark or registered trademark of Epic Games, Inc. in the United States of America and elsewhere.

Abstract Developing robust drone detection systems is often constrained by the limited availability of large-scale annotated training data and the high costs associated with real-world data collection. However, synthetic data presents a promising and cost-effective solution to overcome this issue. Therefore, we present SynDroneVision, a synthetic dataset specifically designed for RGB-based drone detection in surveillance applications. Featuring diverse backgrounds, lighting conditions, and drone models, SynDroneVision offers a comprehensive training foundation for deep learning algorithms. To evaluate the dataset's effectiveness, we perform a comparative analysis across a selection of recent YOLO detection models. Our findings demonstrated that SynDroneVision is a valuable resource for real-world data enrichment, achieving notable enhancements in model performance and robustness, while significantly reducing the time and costs of real-world data acquisition.  PaperPublished in the Proceedings of the 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV2025)!SynDroneVision is presented in the paper SynDroneVision: A Synthetic Dataset for Image-Based Drone Detection by Tamara R. Lenhard, Andreas Weinmann, Kai Franke, and Tobias Koch. This work is published in the Proceedings of the 2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV2025).The preprint is currently available on ArXiv: hereThe final version is now published in the proceedings of WACV 2025: here Dataset DetailsSynDroneVision comprises a total of 140,038 annotaed RGB images (131,238 for training, 8,800 for validation, and 4,000 for testing), featuring a resolution of 2560x1489 pixels. All images are recorded in a sequential manner using Unreal Engine 5.0 in combination with Colosseum. Apart from drone images, SynDroneVision also includes ~7% of background images (i.e., imag frames without drone instances).Annotation Format: Annotations (bounding boxes) are provided via text files according to the YOLO standard format Here, and represent the normalized coordinates of the bounding box center, while and denote the normalized bounding box wisth and height. In SynDroneVision, is always 0, indicating the drone class.DownloadThe SynDroneVision dataset offers around 900 GB of data dedicated to image-based drone detection. To facilitate the download process, we have partitioned the dataset into smaller sections. Specifically, we have divided the training data into 10 segments, organized by sequences.Annotations are available below, with image data accessible via the following links:Dataset SplitSequencesFile NameLinkSize (GB)Training SetSeq. 001 - 009images_train_seq001-009.zipTraining images PART 157 Seq. 010 - 018images_train_seq010-018.zipTrainng images PART 295.4 Seq. 019 - 027images_train_seq019-027.zipTraining images PART 396.2 Seq. 028 - 035images_train_seq028-035.zipTraining images PART 483.9 Seq. 036 - 043images_train_seq036-043.zipTraining images PART 577.1 Seq. 044 - 050images_train_seq044-050.zipTraining images PART 684.7 Seq. 051 - 056images_train_seq051-056.zipTraining images PART 786.8 Seq. 057 - 065images_train_seq057-065.zipTraining images PART 886.2 Seq. 066 - 070images_train_seq066-070.zipTraining images PART 975.7 Seq. 071 - 073images_train_seq071-073.zipTraining images PART 1038.5Validation SetSeq. 001 - 073images_val.zipValidation images55.2Test SetSeq. 001 - 073images_test.zipTest images26.5CitationIf you find SynDroneVision helpful in your research, we kindly ask that you cite the associated paper. Below is the citation in BibTeX format for your convenience:BibTeX:@INPROCEEDINGS{10943801, author={Lenhard, Tamara R. and Weinmann, Andreas and Franke, Kai and Koch, Tobias}, booktitle={2025 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)}, title={SynDroneVision: A Synthetic Dataset for Image-Based Drone Detection}, year={2025}, volume={}, number={}, pages={7637-7647}, doi={10.1109/WACV61041.2025.00742}}SynDroneVision uses Unreal® Engine. Unreal® is a trademark or registered trademark of Epic Games, Inc. in the United States of America and elsewhere.