UAV Cinematography Constraints Imposed by Visual Target Tracking

Karakostas, Iason; Mademlis, Ioannis; Nikolaidis, Nikos; Pitas, Ioannis

doi:10.1109/icip.2018.8451385

Cited by 20 publications

(5 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…UAVs are becoming increasingly common for civilian use, with the Chinese manufacturer DJI dominating the market. They are used in such fields as surveillance [12], filmmaking [13], scientific research, surveying, disaster relief [14], law enforcement, archaeology [15], forestry and agriculture, as well as recreation. The BGS, like many other organizations, are able to utilize the usability and portability of these aircraft to create photogrammetric images and point clouds using SfM techniques, at a fraction of the cost of using a system mounted on an airplane.…”

Section: Geodetic Technologymentioning

confidence: 99%

The Application of Terrestrial LiDAR for Geohazard Mapping, Monitoring and Modelling in the British Geological Survey

Jones

Hobbs

2021

Remote Sensing

View full text Add to dashboard Cite

Geomatics is the discipline of electronically gathering, storing, processing, and delivering spatially related digital information; it continues to be one of the fastest expanding global markets, driven by technology. The British Geological Survey (BGS) geomatics capabilities have been utilized in a variety of scientific studies such as the monitoring of actively growing volcanic lava domes and rapidly retreating glaciers; coastal erosion and platform evolution; inland and coastal landslide modelling; mapping of geological structures and fault boundaries; rock stability and subsidence feature analysis, and geo-conservation. In 2000, the BGS became the first organization outside the mining industry to use Terrestrial LiDAR Scanning (TLS) as a tool for measuring change; paired with a Global Navigation Satellite System (GNSS), BGS were able to measure, monitor, and model geomorphological features of landslides in the United Kingdom (UK) digitally. Many technologies are used by the BGS to monitor the earth, employed on satellites, airplanes, drones, and ground-based equipment, in both research and commercial settings to carry out mapping, monitoring, and modelling of earth surfaces and processes. Outside BGS, these technologies are used for close-range, high-accuracy applications such as bridge and dam monitoring, crime and accident scene analysis, forest canopy and biomass measurements and military applications.

show abstract

Section: Geodetic Technologymentioning

confidence: 99%

The Application of Terrestrial LiDAR for Geohazard Mapping, Monitoring and Modelling in the British Geological Survey

Jones

Hobbs

2021

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Thus, weights w i encoding the degree to which we take each detector's decision into account regarding the currently processed voxel are denoted in Eq. (2). When the detector outputs (either positive or negative ones) are confident, this weight should be higher.…”

Section: Bayesian Fusionmentioning

confidence: 99%

Multiview vision-based human crowd localization for UAV fleet flight safety

Kakaletsis

Mademlis

Nikolaidis

et al. 2021

Signal Processing: Image Communication

Self Cite

View full text Add to dashboard Cite

“…According to recent UAV shot type taxonomy proposals [3], [4], [5], [6], [7], [8], UAV shots consist in valid combinations This work has received funding from the European Union's Horizon 2020 programme under grant agreement No 951911 (AI4Media).…”

Section: Introductionmentioning

confidence: 99%

Deep Reinforcement Learning with semi-expert distillation for autonomous UAV cinematography

Sochopoulos,

Mademlis,

Charalampakis

et al. 2023

2023 IEEE International Conference on Multimedia and Expo (ICME)

Self Cite

View full text Add to dashboard Cite

Unmanned Aerial Vehicles (UAVs, or drones) have revolutionized modern media production. Being rapidly deployable "flying cameras", they can easily capture aesthetically pleasing aerial footage of static or moving filming targets/subjects. Current approaches rely either on manual UAV/gimbal control by human experts, or on a combination of complex computer vision algorithms and hardware configurations for automating the flight+filming process. This paper explores an efficient Deep Reinforcement Learning (DRL) alternative, which implicitly merges the target detection and path planning steps into a single algorithm. To achieve this, a baseline DRL approach is augmented with a novel policy distillation component, which transfers knowledge from a suitable, semi-expert Model Predictive Control (MPC) controller into the DRL agent. Thus, the latter is able to autonomously execute a specific UAV cinematography task with purely visual input. Unlike the MPC controller, the proposed DRL agent does not need to know the 3D world position of the filming target during inference. Experiments conducted in a photorealistic simulator showcase superior performance and training speed compared to the baseline agent, while surpassing the MPC controller in terms of visual occlusion avoidance.

show abstract

UAV Cinematography Constraints Imposed by Visual Target Tracking

Cited by 20 publications

References 8 publications

The Application of Terrestrial LiDAR for Geohazard Mapping, Monitoring and Modelling in the British Geological Survey

The Application of Terrestrial LiDAR for Geohazard Mapping, Monitoring and Modelling in the British Geological Survey

Multiview vision-based human crowd localization for UAV fleet flight safety

Deep Reinforcement Learning with semi-expert distillation for autonomous UAV cinematography

Contact Info

Product

Resources

About