ThermalSynth: A Novel Approach for Generating Synthetic Thermal Human Scenarios

Madan, Neelu; Siemon, Mia; Gjerde, Magnus Kaufmann; Petersson, Bastian Starup; Grotuzas, Arijus; Esbensen, Malthe Aaholm; Nikolov, Ivan Adriyanov; Philipsen, Mark Philip; Nasrollahi, Kamal; Moeslund, Thomas B.

doi:10.1109/wacvw58289.2023.00018

Cited by 3 publications

(2 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Augmenting real images and videos with synthetic elements, on the other hand, bridges the gap much better between real and generated data, as only certain elements of the images are synthesized. Especially for surveillance purposes, such augmentations have been shown to produce good results for pedestrian anomalies [21], falling in water detection [24], camera tampering [22], and subtle anomalies such as dropping objects and animals in the frame, as seen in the work that the current paper is an extension of [20].…”

Section: Synthetic Datasetsmentioning

confidence: 99%

“…Otherwise, they can be created through the augmentation of already existing real-world images with synthetic elements [20][21][22]. While most synthetic data augmentation relies on time-consuming processes, with weeks being spent on frame rendering in some cases [21,23], others, such as the works on responsive anomaly generation [20] and fall synthesis [24], are using real-time rendering game engines to speed up and streamline the process. Even these applications have the downside of still relying on domain-specific knowledge of game developers and 3D artists to set up the environments and generate scenario variations.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

User-Centered Pipeline for Synthetic Augmentation of Anomaly Detection Datasets

Rosbak-Mortensen,

Jansen,

Muhlig

et al. 2024

Computers

View full text Add to dashboard Cite

Automatic anomaly detection plays a critical role in surveillance systems but requires datasets comprising large amounts of annotated data to train and evaluate models. Gathering and annotating these data is a labor-intensive task that can become costly. A way to circumvent this is to use synthetic data to augment anomalies directly into existing datasets. This far more diverse scenario can be created and come directly with annotations. This however also poses new issues for the computer-vision engineer and researcher end users, who are not readily familiar with 3D modeling, game development, or computer graphics methodologies and must rely on external specialists to use or tweak such pipelines. In this paper, we extend our previous work of an application that synthesizes dataset variations using 3D models and augments anomalies on real backgrounds using the Unity Engine. We developed a high-usability user interface for our application through a series of RITE experiments and evaluated the final product with the help of deep-learning specialists who provided positive feedback regarding its usability, accessibility, and user experience. Finally, we tested if the proposed solution can be used in the context of traffic surveillance by augmenting the train data from the challenging Street Scene dataset. We found that by using our synthetic data, we could achieve higher detection accuracy. We also propose the next steps to expand the proposed solution for better usability and render accuracy through the use of segmentation pre-processing.

show abstract

Section: Synthetic Datasetsmentioning

confidence: 99%