arXiv subjects

John McDonald

Publications and source records attributed to John McDonald.

At least 19 recordsLinked to original sources

arXiv2025

Benchmarking SAM2-based Trackers on FMOX

Several object tracking pipelines extending Segment Anything Model 2 (SAM2) have been proposed in the past year, where the approach is to follow and segment the object from a single exemplar template provided by the user on a initialization frame. We propose to benchmark these high performing trackers (SAM2, EfficientTAM, DAM4SAM and SAMURAI) on datasets containing fast moving objects (FMO) specifically designed to be challenging for tracking approaches. The goal is to understand better current limitations in state-of-the-art trackers by providing more detailed insights on the behavior of these trackers. We show that overall the trackers DAM4SAM and SAMURAI perform well on more challenging sequences.

cs.CV↗

arXiv2025

Benchmarking EfficientTAM on FMO datasets

Fast and tiny object tracking remains a challenge in computer vision and in this paper we first introduce a JSON metadata file associated with four open source datasets of Fast Moving Objects (FMOs) image sequences. In addition, we extend the description of the FMOs datasets with additional ground truth information in JSON format (called FMOX) with object size information. Finally we use our FMOX file to test a recently proposed foundational model for tracking (called EfficientTAM) showing that its performance compares well with the pipelines originally taylored for these FMO datasets. Our comparison of these state-of-the-art techniques on FMOX is provided with Trajectory Intersection of Union (TIoU) scores. The code and JSON is shared open source allowing FMOX to be accessible and usable for other machine learning pipelines aiming to process FMO datasets.

cs.CV↗

arXiv2025

Higgs Inflation with Vector-Like Quark Stabilisation and the ACT spectral index

Recently, the Atacama Cosmology Telescope (ACT) collaboration has reported a scalar spectral index $n_s~=~0.9743~\pm~0.0034$. This is substantially larger than the classical prediction of Higgs Inflation, $n_s \approx 0.965$, which is 2.74$\sigma$ below the ACT mean value. We show that when an otherwise metastable Standard Model Higgs Inflation potential is stabilised by the addition of vector-like quark pairs and the potential is renormalised in the Jordan frame, the value of $n_s$ is generally larger than 0.965 and can explain the ACT observation. As an example, assuming the 2022 PDG direct measurement central value for the t quark mass, $m_{t} = 172.69$ GeV, and central values for the SM inputs to the renormalisation group equations, we obtain $n_s = 0.9792 - 0.9844$ for the case of three isosinglet vector-like B quarks with mass $m_{Q}$ in the range 1-3 TeV, with the lowest value of the $n_s$ range being 1.44$\sigma$ above the ACT mean value. The model predicts primordial gravitational wave with tensor-to-scalar ratio $r = 7.87 \times 10^{-3} - 1.21 \times 10^{-2}$ for $m_{Q} =$ 1-3 TeV, which will be easily observable in forthcoming CMB experiments. Observation of vector-like quarks of mass close to 1 TeV mass combined with a large tensor-to-scalar ratio $r \sim 0.01$ would support the model.

hep-ph↗

arXiv2025

Vector-like Quark Stabilised Higgs Inflation: Implications for Particle Phenomenology, Primordial Gravitational Waves and the Hubble Tension

The Standard Model (SM) Higgs potential is likely to be metastable, in which case Higgs Inflation requires an extension of the SM to sufficiently stabilise the Higgs potential. Here we consider stabilisation by adding $n_{Q} \leq 3$ Vector-Like Quarks (VLQs) of mass $m_{Q}$. We consider isosinglet $T$ and $B$ vector quarks. Requiring stability of the finite temperature effective potential, we find that the upper bounds on $m_{Q}$ for $T$ quarks are 5.8 TeV (for $n_{Q} = 2$) and 55 TeV (for $n_{Q} = 3$). The upper bounds are generally smaller for $B$ vector quarks and are sensitive to the $t$-quark mass. The inflation predictions depend upon the conformal frame in which the model is renormalised. For renormalisation in the Einstein frame (Prescription I) the predictions are almost indistinguishable from the classical values: $n_s = 0.966$ and $r = 3.3 \times 10^{-3}$. Renormalisation in the Jordan frame (Prescription II) predicts larger values of $n_{s}$ and $r$, with $n_{s}$ generally in the range 0.980 to 0.990 and $r$ of the order of 0.01. The predicted range of $n_{s}$ is consistent with the CMB range obtained in Hubble tension solutions which modify the sound horizon at decoupling, whilst the predicted values of $r$ will be easily observable by forthcoming CMB experiments. The observational upper bound on $r$ generally imposes a stronger upper bound on $m_{Q}$ in Prescription II than the requirement of stability. We conclude that VLQ-stabilised Higgs Inflation with Prescription II renormalisation favours 1-10 TeV vector-like quarks that will be accessible to future colliders, and predicts a tensor-to-scalar ratio that will be observable in forthcoming CMB experiments and values of $n_{s}$ that favour an early-time solution to the Hubble tension.

hep-ph↗

arXiv2024

Scalable and Efficient Hierarchical Visual Topological Mapping

Hierarchical topological representations can significantly reduce search times within mapping and localization algorithms. Although recent research has shown the potential for such approaches, limited consideration has been given to the suitability and comparative performance of different global feature representations within this context. In this work, we evaluate state-of-the-art hand-crafted and learned global descriptors using a hierarchical topological mapping technique on benchmark datasets and present results of a comprehensive evaluation of the impact of the global descriptor used. Although learned descriptors have been incorporated into place recognition methods to improve retrieval accuracy and enhance overall recall, the problem of scalability and efficiency when applied to longer trajectories has not been adequately addressed in a majority of research studies. Based on our empirical analysis of multiple runs, we identify that continuity and distinctiveness are crucial characteristics for an optimal global descriptor that enable efficient and scalable hierarchical mapping, and present a methodology for quantifying and contrasting these characteristics across different global descriptors. Our study demonstrates that the use of global descriptors based on an unsupervised learned Variational Autoencoder (VAE) excels in these characteristics and achieves significantly lower runtime. It runs on a consumer grade desktop, up to 2.3x faster than the second best global descriptor, NetVLAD, and up to 9.5x faster than the hand-crafted descriptor, PHOG, on the longest track evaluated (St Lucia, 17.6 km), without sacrificing overall recall performance.

cs.CV↗

arXiv2024

WoodScape Motion Segmentation for Autonomous Driving -- CVPR 2023 OmniCV Workshop Challenge

Motion segmentation is a complex yet indispensable task in autonomous driving. The challenges introduced by the ego-motion of the cameras, radial distortion in fisheye lenses, and the need for temporal consistency make the task more complicated, rendering traditional and standard Convolutional Neural Network (CNN) approaches less effective. The consequent laborious data labeling, representation of diverse and uncommon scenarios, and extensive data capture requirements underscore the imperative of synthetic data for improving machine learning model performance. To this end, we employ the PD-WoodScape synthetic dataset developed by Parallel Domain, alongside the WoodScape fisheye dataset. Thus, we present the WoodScape fisheye motion segmentation challenge for autonomous driving, held as part of the CVPR 2023 Workshop on Omnidirectional Computer Vision (OmniCV). As one of the first competitions focused on fisheye motion segmentation, we aim to explore and evaluate the potential and impact of utilizing synthetic data in this domain. In this paper, we provide a detailed analysis on the competition which attracted the participation of 112 global teams and a total of 234 submissions. This study delineates the complexities inherent in the task of motion segmentation, emphasizes the significance of fisheye datasets, articulate the necessity for synthetic datasets and the resultant domain gap they engender, outlining the foundational blueprint for devising successful solutions. Subsequently, we delve into the details of the baseline experiments and winning methods evaluating their qualitative and quantitative results, providing with useful insights.

cs.CV↗

arXiv2023

GP-net: Flexible Viewpoint Grasp Proposal

We present the Grasp Proposal Network (GP-net), a Convolutional Neural Network model which can generate 6-DoF grasps from flexible viewpoints, e.g. as experienced by mobile manipulators. To train GP-net, we synthetically generate a dataset containing depth-images and ground-truth grasp information. In real-world experiments, we use the EGAD evaluation benchmark to evaluate GP-net against two commonly used algorithms, the Volumetric Grasping Network (VGN) and the Grasp Pose Detection package (GPD), on a PAL TIAGo mobile manipulator. In contrast to the state-of-the-art methods in robotic grasping, GP-net can be used for grasping objects from flexible, unknown viewpoints without the need to define the workspace and achieves a grasp success of 54.4% compared to 51.6% for VGN and 44.2% for GPD. We provide a ROS package along with our code and pre-trained models at https://aucoroboticsmu.github.io/GP-net/.

cs.RO↗

arXiv2023

Higgs Inflation Using the Unstable Standard Model Potential

It is likely that the Higgs potential of the Standard Model is unstable, turning negative at $ϕ< Λ\sim 10^{10}$ GeV. Here we consider whether it is possible to have Higgs Inflation on the positive stable region of the potential at $ϕ< Λ$. To do this we add a non-minimally coupled induced gravity sector with scalar $χ$ to the Standard Model. For an appropriate form for the non-minimal coupling of $χ$, we show that it is possible to have conventional Higgs inflation at small $ϕ< Λ$ if the effective Planck mass in the Jordan frame during inflation is sufficiently small, with a phase transition to $χ\neq 0$ at the end of Higgs inflation which increases the Jordan frame Planck mass to its presently observed value. In the Einstein frame this corresponds to a suppression of the Higgs kinetic and potential term at the end of inflation. We show that the predictions of Higgs inflation at tree level are unaltered from conventional Higgs Inflation, with the exception of the magnitude of the Higgs field during inflation. Hence Higgs Inflation can be achieved using the potential of the unmodified Standard Model.

hep-ph↗

arXiv2023

Leptogenesis via Inflaton Mass Terms in Non-Minimally Coupled Inflation

We consider a model of baryogenesis based on adding lepton number-violating quadratic mass terms to the inflaton potential of a non-minimally coupled inflation model. The $L$-violating mass terms generate a lepton asymmetry in a complex inflaton field via the mass term Affleck-Dine mechanism, which is transferred to the Standard Model (SM) sector when the inflaton decays to right-handed (RH) neutrinos. The model is minimal in that it requires only the SM sector, RH neutrinos, and a non-minimally coupled inflaton sector. We find that baryon isocurvature fluctuations can be observable in metric inflation but are negligible in Palatini inflation. The model is compatible with reheating temperatures that may be detectable in the observable primordial gravitational waves predicted by metric inflation.

hep-ph↗

arXiv2022

Fast and Efficient Scene Categorization for Autonomous Driving using VAEs

Scene categorization is a useful precursor task that provides prior knowledge for many advanced computer vision tasks with a broad range of applications in content-based image indexing and retrieval systems. Despite the success of data driven approaches in the field of computer vision such as object detection, semantic segmentation, etc., their application in learning high-level features for scene recognition has not achieved the same level of success. We propose to generate a fast and efficient intermediate interpretable generalized global descriptor that captures coarse features from the image and use a classification head to map the descriptors to 3 scene categories: Rural, Urban and Suburban. We train a Variational Autoencoder in an unsupervised manner and map images to a constrained multi-dimensional latent space and use the latent vectors as compact embeddings that serve as global descriptors for images. The experimental results evidence that the VAE latent vectors capture coarse information from the image, supporting their usage as global descriptors. The proposed global descriptor is very compact with an embedding length of 128, significantly faster to compute, and is robust to seasonal and illuminational changes, while capturing sufficient scene information required for scene categorization.

cs.CV↗

arXiv2022

Woodscape Fisheye Object Detection for Autonomous Driving -- CVPR 2022 OmniCV Workshop Challenge

Object detection is a comprehensively studied problem in autonomous driving. However, it has been relatively less explored in the case of fisheye cameras. The strong radial distortion breaks the translation invariance inductive bias of Convolutional Neural Networks. Thus, we present the WoodScape fisheye object detection challenge for autonomous driving which was held as part of the CVPR 2022 Workshop on Omnidirectional Computer Vision (OmniCV). This is one of the first competitions focused on fisheye camera object detection. We encouraged the participants to design models which work natively on fisheye images without rectification. We used CodaLab to host the competition based on the publicly available WoodScape fisheye dataset. In this paper, we provide a detailed analysis on the competition which attracted the participation of 120 global teams and a total of 1492 submissions. We briefly discuss the details of the winning methods and analyze their qualitative and quantitative results.

cs.CV↗

arXiv2022

ViT-BEVSeg: A Hierarchical Transformer Network for Monocular Birds-Eye-View Segmentation

Generating a detailed near-field perceptual model of the environment is an important and challenging problem in both self-driving vehicles and autonomous mobile robotics. A Bird Eye View (BEV) map, providing a panoptic representation, is a commonly used approach that provides a simplified 2D representation of the vehicle surroundings with accurate semantic level segmentation for many downstream tasks. Current state-of-the art approaches to generate BEV-maps employ a Convolutional Neural Network (CNN) backbone to create feature-maps which are passed through a spatial transformer to project the derived features onto the BEV coordinate frame. In this paper, we evaluate the use of vision transformers (ViT) as a backbone architecture to generate BEV maps. Our network architecture, ViT-BEVSeg, employs standard vision transformers to generate a multi-scale representation of the input image. The resulting representation is then provided as an input to a spatial transformer decoder module which outputs segmentation maps in the BEV grid. We evaluate our approach on the nuScenes dataset demonstrating a considerable improvement in the performance relative to state-of-the-art approaches.

cs.CV↗

arXiv2022

General Analytical Conditions for Inflaton Fragmentation: Quick and Easy Tests for its Occurrence

Understanding the physics of inflaton condensate fragmentation in the early Universe is crucial as the existence of fragments in the form of non-topological solitons (oscillons or Q-balls) may potentially modify the evolution of the post-inflation Universe. Furthermore, such fragments may evolve into primordial black holes and form dark matter, or emit gravitational waves. Due to the non-perturbative and non-linear nature of the dynamics, most of the studies rely on numerical lattice simulations. Numerical simulations of condensate fragmentation are, however, challenging and, without knowing where to look in the parameter space, they are likely to be time-consuming as well. In this paper, we provide generic analytical conditions for the perturbations of an inflaton condensate to undergo growth to non-linearity in the cases of both symmetric and asymmetric inflaton potentials. We apply the conditions to various inflation models and demonstrate that our results are in good agreement with explicit numerical simulations. Our analytical conditions are easy to use and may be utilised in order to quickly identify models that may undergo fragmentation and determine the conditions under which they do so, which can guide subsequent in-depth numerical analyses.

astro-ph.CO↗

arXiv2021

FisheyeSuperPoint: Keypoint Detection and Description Network for Fisheye Images

Keypoint detection and description is a commonly used building block in computer vision systems particularly for robotics and autonomous driving. However, the majority of techniques to date have focused on standard cameras with little consideration given to fisheye cameras which are commonly used in urban driving and automated parking. In this paper, we propose a novel training and evaluation pipeline for fisheye images. We make use of SuperPoint as our baseline which is a self-supervised keypoint detector and descriptor that has achieved state-of-the-art results on homography estimation. We introduce a fisheye adaptation pipeline to enable training on undistorted fisheye images. We evaluate the performance on the HPatches benchmark, and, by introducing a fisheye based evaluation method for detection repeatability and descriptor matching correctness, on the Oxford RobotCar dataset.

cs.CV↗

arXiv2021

2.5D Vehicle Odometry Estimation

It is well understood that in ADAS applications, a good estimate of the pose of the vehicle is required. This paper proposes a metaphorically named 2.5D odometry, whereby the planar odometry derived from the yaw rate sensor and four wheel speed sensors is augmented by a linear model of suspension. While the core of the planar odometry is a yaw rate model that is already understood in the literature, we augment this by fitting a quadratic to the incoming signals, enabling interpolation, extrapolation, and a finer integration of the vehicle position. We show, by experimental results with a DGPS/IMU reference, that this model provides highly accurate odometry estimates, compared with existing methods. Utilising sensors that return the change in height of vehicle reference points with changing suspension configurations, we define a planar model of the vehicle suspension, thus augmenting the odometry model. We present an experimental framework and evaluations criteria by which the goodness of the odometry is evaluated and compared with existing methods. This odometry model has been designed to support low-speed surround-view camera systems that are well-known. Thus, we present some application results that show a performance boost for viewing and computer vision applications using the proposed odometry

cs.RO↗

arXiv2021

Woodscape Fisheye Semantic Segmentation for Autonomous Driving -- CVPR 2021 OmniCV Workshop Challenge

We present the WoodScape fisheye semantic segmentation challenge for autonomous driving which was held as part of the CVPR 2021 Workshop on Omnidirectional Computer Vision (OmniCV). This challenge is one of the first opportunities for the research community to evaluate the semantic segmentation techniques targeted for fisheye camera perception. Due to strong radial distortion standard models don't generalize well to fisheye images and hence the deformations in the visual appearance of objects and entities needs to be encoded implicitly or as explicit knowledge. This challenge served as a medium to investigate the challenges and new methodologies to handle the complexities with perception on fisheye images. The challenge was hosted on CodaLab and used the recently released WoodScape dataset comprising of 10k samples. In this paper, we provide a summary of the competition which attracted the participation of 71 global teams and a total of 395 submissions. The top teams recorded significantly improved mean IoU and accuracy scores over the baseline PSPNet with ResNet-50 backbone. We summarize the methods of winning algorithms and analyze the failure cases. We conclude by providing future directions for the research.

cs.CV↗

arXiv2021

OdoViz: A 3D Odometry Visualization and Processing Tool

OdoViz is a reactive web-based tool for 3D visualization and processing of autonomous vehicle datasets designed to support common tasks in visual place recognition research. The system includes functionality for loading, inspecting, visualizing, and processing GPS/INS poses, point clouds and camera images. It supports a number of commonly used driving datasets and can be adapted to load custom datasets with minimal effort. OdoViz's design consists of a slim server to serve the datasets coupled with a rich client frontend. This design supports multiple deployment configurations including single user stand-alone installations, research group installations serving datasets internally across a lab, or publicly accessible web-frontends for providing online interfaces for exploring and interacting with datasets. The tool allows viewing complete vehicle trajectories traversed at multiple different time periods simultaneously, facilitating tasks such as sub-sampling, comparing and finding pose correspondences both across and within sequences. This significantly reduces the effort required in creating subsets of data from existing datasets for machine learning tasks. Further to the above, the system also supports adding custom extensions and plugins to extend the capabilities of the software for other potential data management, visualization and processing tasks. The platform has been open-sourced to promote its use and encourage further contributions from the research community.

cs.CV↗

arXiv2021

A Minimal Approach to Baryogenesis via Affleck-Dine and Inflaton Mass Terms

We present a minimal approach to the generation of the baryon ($B$) asymmetry of the Universe, in which the asymmetry is generated in a complex inflaton condensate via $B$-violating quadratic inflaton potential terms using the Affleck-Dine (AD) mechanism. We show that the $B$-violating quadratic mass terms create an oscillating asymmetry in the complex inflaton condensate at late times. The final asymmetry transferred to the Standard Model sector at reheating is naturally reduced to the magnitude of the observed $B$ asymmetry by the effect of averaging over the $B$ oscillations. This approach to baryogenesis can easily be realised in a wide range of inflation models. We also show that AD baryogenesis via inflaton mass terms is free of the tuning of field dynamics required by the conventional AD mechanism.

hep-ph↗