70 research outputs found
Dataset: Mask R-CNN Based C. Elegans Detection with a DIY Microscope
No full textThe dataset consists of images of C. elegans in Petri Dish that were captured at a frequency of 1 Hz at 3280 × 2464 pixels via a Raspberry Pi based DIY Microscope. Further details of the recording setup and the dataset can be found in the corresponding article.
Up on use, please cite the following article https://doi.org/10.3390/bios11080257 such as:
Fudickar, S.; Nustede, E.J.; Dreyer, E.; Bornhorst, J. Mask R-CNN Based C. Elegans Detection with a DIY Microscope. Biosensors 2021, 11, 257. https://doi.org/10.3390/bios11080257Usage:
To train the neural network, the specified folder for the RGB images is checked first with the os library and the paths of all contained jpg files are stored as strings in a list.
This list is then iterated and the first image is loaded via the openCV library.
The filename of each RGB image contains a number and an ID as unique identifier. This is used to identify the associated labels. For example, image with Filename "img_038_id2.jpg" has the ID 2.
In the associated masks folder, all image files are loaded that also have the 038_id2 in the file name, such as "m5_img_038_id2.jpg". So for each RGB image a list of mask images, also loaded with openCV, is kept in temporary memory. All images then run through the sliding window algorithm and are presented to the network in individual parts one after the other
Estimation of Psychosocial Work Environment Exposures Through Video Object Detection:Proof of Concept Using CCTV Footage
No full textThis paper examines the use of computer vision algorithms to estimate aspects of the psychosocial work environment using CCTV footage. We present a proof of concept for a methodology that detects and tracks people in video footage and estimates interactions between customers and employees by estimating their poses and calculating the duration of their encounters. We propose a pipeline that combines existing object detection and tracking algorithms (YOLOv8 and DeepSORT) with pose estimation algorithms (BlazePose) to estimate the number of customers and employees in the footage as well as the duration of their encounters. We use a simple rule-based approach to classify the interactions as positive, neutral or negative based on three different criteria: distance, duration and pose. The proposed methodology is tested on a small dataset of CCTV footage. While the data is quite limited in particular with respect to the quality of the footage, we have chosen this case as it represents a typical setting where the method could be applied. The results show that the object detection and tracking part of the pipeline has a reasonable performance on the dataset with a high degree of recall and reasonable accuracy. At this stage, the pose estimation is still limited to fully detect the type of interactions due to difficulties in tracking employees in the footage. We conclude that the method is a promising alternative to self-reported measures of the psychosocial work environment and could be used in future studies to obtain external observations of the work environment.This paper examines the use of computer vision algorithms to estimate aspects of the psychosocial work environment using CCTV footage. We present a proof of concept for a methodology that detects and tracks people in video footage and estimates interactions between customers and employees by estimating their poses and calculating the duration of their encounters. We propose a pipeline that combines existing object detection and tracking algorithms (YOLOv8 and DeepSORT) with pose estimation algorithms (BlazePose) to estimate the number of customers and employees in the footage as well as the duration of their encounters. We use a simple rule-based approach to classify the interactions as positive, neutral or negative based on three different criteria: distance, duration and pose. The proposed methodology is tested on a small dataset of CCTV footage. While the data is quite limited in particular with respect to the quality of the footage, we have chosen this case as it represents a typical setting where the method could be applied. The results show that the object detection and tracking part of the pipeline has a reasonable performance on the dataset with a high degree of recall and reasonable accuracy. At this stage, the pose estimation is still limited to fully detect the type of interactions due to difficulties in tracking employees in the footage. We conclude that the method is a promising alternative to self-reported measures of the psychosocial work environment and could be used in future studies to obtain external observations of the work environment
Sensor-Based Activity Recognition and Artificial Intelligence
No full textThis book constitutes the proceedings of the 9th International Workshop on Sensor-Based Activity Recognition and Artificial Intelligence, iWOAR 2024, held in Potsdam, Germany, during September 26–27, 2024. The 15 full papers and 4 short papers presented here were carefully reviewed and selected from 28 submissions. These papers have been categorized into the following topical sections: Advances in Human Activity Recognition; Applications in Vision-Based Recognition; Wearable Devices and Health Monitoring; Novel AI and Machine Learning Approaches; Short Papers: Emerging Topics in Sensor-Based Systems
Comparing suitability of sub 1 GHz and WiFi transceivers for RSS-based indoor localisation
No full text
TMNet - Distributed viewing and editing of Topic Maps in the World Wide Web Environment
No full textSince the Topic Map standard describes a prospective knowledge-structuring model that can be used in a huge variety of knowledge domains the amount of applications utilizing this standard grew enormously. Anyhow, as far as we know, there is no distributed editor of Topic Maps available on the market, that supports an intuitive possibility to manipulate or visualize Topic Maps in the World Wide Web environment. Therefore, TMNet intends to close this gap. As a result, a high performing distributed editor is achieved, with which web-based e-learning environments as well as knowledge environments can be enhanced by supporting semantic data descriptions of their knowledge structure. Additionally TMNet enables the user to extend the structures in an easy way and gives the web developer an opportunity to integrate it into websites easily
- …
