arXiv subjects

Andrew Price

Publications and source records attributed to Andrew Price.

8 recordsLinked to original sources

arXiv2026

Streak detection in the VST/OmegaCAM archive using deep learning

Ground-based astronomical surveys inadvertently capture streaks from satellites and space debris crossing their fields of view. These incidental observations from wide-field instruments such as OmegaCAM on the VST offer valuable opportunities to characterise resident space objects without the need for dedicated observing time. We developed an automated deep-learning pipeline to detect and classify streaks in the OmegaCAM archive, enabling large-scale analyses of space object populations and their impact on astronomical data. The pipeline combines an adapted Hough transform lookup-based convolutional neural network (HT-LCNN) for initial streak detection on raw images with a VGG6-based CNN classifier to reject false positives. We augmented a manually annotated dataset of 384 000 patches from archive images with physically simulated streaks. Following a detection, we applied astrometric calibration and cross-matched the results with the space-track catalogue. We find the detector achieves F1-scores of 0.966 (validation) and 0.958 (test) on the augmented dataset, detecting > 95% of artificial streaks with a signal-to-noise ratio of S/N > 4. On real 2023 data, the precision drops to 0.783 due to image variability, but the classifier boosts it to 0.990, while retaining 97% of true positives and rejecting > 96% of false positives. Applied to one year of VST observations (1 246 048 OmegaCAM CCD frames), the pipeline identified 25 335 streaks, including more than 20% uncorrelated with catalogue entries; finally, 16.9% of images revealed some level of contamination. The pipeline demonstrates robust performance on real archival data and successfully uncovers faint uncatalogued objects, highlighting the potential of survey archives for debris monitoring.

astro-ph.IM↗

arXiv2025

FastPose-ViT: A Vision Transformer for Real-Time Spacecraft Pose Estimation

Estimating the 6-degrees-of-freedom (6DoF) pose of a spacecraft from a single image is critical for autonomous operations like in-orbit servicing and space debris removal. Existing state-of-the-art methods often rely on iterative Perspective-n-Point (PnP)-based algorithms, which are computationally intensive and ill-suited for real-time deployment on resource-constrained edge devices. To overcome these limitations, we propose FastPose-ViT, a Vision Transformer (ViT)-based architecture that directly regresses the 6DoF pose. Our approach processes cropped images from object bounding boxes and introduces a novel mathematical formalism to map these localized predictions back to the full-image scale. This formalism is derived from the principles of projective geometry and the concept of "apparent rotation", where the model predicts an apparent rotation matrix that is then corrected to find the true orientation. We demonstrate that our method outperforms other non-PnP strategies and achieves performance competitive with state-of-the-art PnP-based techniques on the SPEED dataset. Furthermore, we validate our model's suitability for real-world space missions by quantizing it and deploying it on power-constrained edge hardware. On the NVIDIA Jetson Orin Nano, our end-to-end pipeline achieves a latency of ~75 ms per frame under sequential execution, and a non-blocking throughput of up to 33 FPS when stages are scheduled concurrently.

cs.CV↗

arXiv2023

Modular Quantization-Aware Training for 6D Object Pose Estimation

Edge applications, such as collaborative robotics and spacecraft rendezvous, demand efficient 6D object pose estimation on resource-constrained embedded platforms. Existing 6D pose estimation networks are often too large for such deployments, necessitating compression while maintaining reliable performance. To address this challenge, we introduce Modular Quantization-Aware Training (MQAT), an adaptive and mixed-precision quantization-aware training strategy that exploits the modular structure of modern 6D pose estimation architectures. MQAT guides a systematic gradated modular quantization sequence and determines module-specific bit precisions, leading to quantized models that outperform those produced by state-of-the-art uniform and mixed-precision quantization techniques. Our experiments showcase the generality of MQAT across datasets, architectures, and quantization algorithms. Remarkably, MQAT-trained quantized models achieve a significant accuracy boost (>7%) over the baseline full-precision network while reducing model size by a factor of 4x or more. Our project website is at: https://saqibjaved1.github.io/MQAT_/

cs.CV↗

arXiv2021

Fusing RGBD Tracking and Segmentation Tree Sampling for Multi-Hypothesis Volumetric Segmentation

Despite rapid progress in scene segmentation in recent years, 3D segmentation methods are still limited when there is severe occlusion. The key challenge is estimating the segment boundaries of (partially) occluded objects, which are inherently ambiguous when considering only a single frame. In this work, we propose Multihypothesis Segmentation Tracking (MST), a novel method for volumetric segmentation in changing scenes, which allows scene ambiguity to be tracked and our estimates to be adjusted over time as we interact with the scene. Two main innovations allow us to tackle this difficult problem: 1) A novel way to sample possible segmentations from a segmentation tree; and 2) A novel approach to fusing tracking results with multiple segmentation estimates. These methods allow MST to track the segmentation state over time and incorporate new information, such as new objects being revealed. We evaluate our method on several cluttered tabletop environments in simulation and reality. Our results show that MST outperforms baselines in all tested scenes.

cs.RO↗

arXiv2020

Using Decision Diagrams to Compactly Represent the State Space for Explicit Model Checking

The enormous number of states reachable during explicit model checking is the main bottleneck for scalability. This paper presents approaches of using decision diagrams to represent very large state space compactly and efficiently. This is possible for asynchronous systems as two system states connected by a transition often share many same local portions. Using decision diagrams can significantly reduce memory demand by not using memory to store the redundant information among different states. This paper considers multi-value decision diagrams for this purpose. Additionally, a technique to reduce the runtime overhead of using these diagrams is also described. Experimental results and comparison with the state compression method as implemented in the model checker SPIN show that the approaches presented in this paper are memory efficient for storing large state space with acceptable runtime overhead.

cs.SE↗

arXiv2019

Inferring Occluded Geometry Improves Performance when Retrieving an Object from Dense Clutter

Object search -- the problem of finding a target object in a cluttered scene -- is essential to solve for many robotics applications in warehouse and household environments. However, cluttered environments entail that objects often occlude one another, making it difficult to segment objects and infer their shapes and properties. Instead of relying on the availability of CAD or other explicit models of scene objects, we augment a manipulation planner for cluttered environments with a state-of-the-art deep neural network for shape completion as well as a volumetric memory system, allowing the robot to reason about what may be contained in occluded areas. We test the system in a variety of tabletop manipulation scenes composed of household items, highlighting its applicability to realistic domains. Our results suggest that incorporating both components into a manipulation planning framework significantly reduces the number of actions needed to find a hidden object in dense clutter.

cs.RO↗

arXiv2013

PyGFit: A Tool for Extracting PSF Matched Photometry

We present PyGFit, a program designed to measure PSF-matched photometry from images with disparate pixel scales and PSF sizes. While PyGFit has a number of uses, its primary purpose is to extract robust spectral energy distributions (SEDs) from crowded images. It does this by fitting blended sources in crowded, low resolution images with models generated from a higher resolution image. This approach minimizes the impact of crowding and also yields consistently measured fluxes in different filters, minimizing systematic uncertainty in the final SEDs. We present an example of applying PyGFit to real data and perform simulations to test its fidelity. The uncertainty in the best-fit flux rises sharply as a function of nearest-neighbor distance for objects with a neighbor within 60% of the PSF size. Similarly, the uncertainty increases quickly for objects blended with a neighbor more than four times brighter. For all other objects the fidelity of PyGFit's results depends only on flux, and the uncertainty is primarily limited by sky noise.

astro-ph.IM↗

arXiv2010

Driving LED in a Nanosecond Regime by a Fast Operational Amplifier

It is widely believed that the generation of high speed optical signals is not the job for an LED. However this work is done to show that there are techniques which can be used to produce nanosecond square pulses from a diode. Rise and fall times of a typical 10ns long signal were 1-2 ns and the intensity of the emission could be controlled by the supply voltage. The wavelength of the radiation was 472 nm, which is blue in colour, but any longer or even shorter wavelengths can similarly be used. The consistency of the experiment and its theoretical model was analysed by computer simulations using OrCAD and PSPICE.

cond-mat.other↗