arXiv subjects

Matthew Watson

Publications and source records attributed to Matthew Watson.

12 recordsLinked to original sources

arXiv2026

Diagnosing Performance in Invasion-Based Esports: The Esports Performance Screening (EPS) Framework

Coaching in esports continues to become more common, yet conceptual tools for coaches to analyse performance breakdowns in esports remain limited. Existing approaches lack a way to distinguish between mental errors (i.e., mistakes relate to strategy and tactics), and physical slips (i.e., motor execution). This paper introduces the Esports Performance Screening (EPS) framework, that integrates several frameworks and models from sport and computing science which can be used to analyse mental and motor performance in esports. The EPS framework organises performance into five interconnected levels: strategy, tactics, tasks, actions and operations. Across these levels, teams pursue forms of superiority that influence transitions between stability and instability during invasion-based esports competition. The framework supports two complementary modes of screening use: diagnostic during reflective review and real-time screening during live play. Through illustrative esports scenarios, we demonstrate how instability can originate at different hierarchical layers and how misattribution of failure can obscure underlying causes. The purpose of the EPS framework is to help coaches systematically diagnose performance breakdowns, and act as heuristics to aid gameplay analysis during a match. The paper contributes a theoretically grounded, title-agnostic esports framework to support systematic performance analysis and pedagogical development in invasion-based esports.

cs.HC↗

arXiv2026

Beyond Shapley: An Influence-Based Data Auditing Pipeline for LLM Alignment and Evaluation

The alignment of Large Language Models (LLMs) is increasingly bottlenecked by data quality. As datasets scale, massive preference and instruction-tuning corpora inevitably accumulate hidden structural contradictions, safety risks, and systemic human annotation errors. Standard dataset auditing methods, such as semantic deduplication or LLM-as-a-judge, struggle to capture the actual predictive impact of individual records and often miss deep functional rule clashes. To address this, we introduce a scalable, inference-only data valuation pipeline that approximates the Shapley value without iterative model retraining. By mapping semantic k-NN neighborhoods into a directed graph, our framework evaluates data utility directly through a reference LLM's probability distribution using zero-shot and one-shot conditional log-likelihood shifts. Our pipeline then translates these predictive influence scores into localized advantage metrics to isolate gradient-conflicting records. We demonstrate the pipeline's efficacy in sanitizing two heavily vetted alignment datasets. First, applying our pipeline to the HelpSteer2 dataset reduced the manual audit search space by 99.1%, successfully uncovering falsely-labeled records across diverse failure modes. Second, applying our automated audit strategy to Anthropic's HH-RLHF training and evaluation splits identified thousands of hidden safety and factual preference inversions. Crucially, by extending this audit to the evaluation split, we expose severe vulnerabilities in current benchmark integrity: highly capable models frequently predict the safer or more helpful response, only to be penalized by objectively flawed human ground-truth labels. Overall, our work provides a mathematically grounded, highly efficient diagnostic tool to uncover human label failures, sanitize evaluation benchmarks, and ensure the integrity of LLM alignment data.

cs.LG↗

arXiv2026

RovoDev Code Reviewer: A Large-Scale Online Evaluation of LLM-based Code Review Automation at Atlassian

Large Language Models (LLMs)-powered code review automation has the potential to transform code review workflows. Despite the advances of LLM-powered code review comment generation approaches, several practical challenges remain for designing enterprise-grade code review automation tools. In particular, this paper aims at answering the practical question: how can we design a review-guided, context-aware, quality-checked code review comment generation without fine-tuning? In this paper, we present RovoDev Code Reviewer, an enterprise-grade LLM-based code review automation tool designed and deployed at scale within Atlassian's development ecosystem with seamless integration into Atlassian's Bitbucket. Through the offline, online, user feedback evaluations over a one-year period, we conclude that RovoDev Code Reviewer is effective in generating code review comments that could lead to code resolution for 38.70% (i.e., comments that triggered code changes in the subsequent commits); and offers the promise of accelerating feedback cycles (i.e., decreasing the PR cycle time by 30.8%), alleviating reviewer workload (i.e., reducing the number of human-written comments by 35.6%), and improving overall software quality (i.e., finding errors with actionable suggestions).

cs.SE↗

arXiv2025

WildLive: Near Real-time Visual Wildlife Tracking onboard UAVs

Live tracking of wildlife via high-resolution video processing directly onboard drones is widely unexplored and most existing solutions rely on streaming video to ground stations to support navigation. Yet, both autonomous animal-reactive flight control beyond visual line of sight and/or mission-specific individual and behaviour recognition tasks rely to some degree on this capability. In response, we introduce WildLive - a near real-time animal detection and tracking framework for high-resolution imagery running directly onboard uncrewed aerial vehicles (UAVs). The system performs multi-animal detection and tracking at 17.81fps for HD and 7.53fps on 4K video streams suitable for operation during higher altitude flights to minimise animal disturbance. Our system is optimised for Jetson Orin AGX onboard hardware. It integrates the efficiency of sparse optical flow tracking and mission-specific sampling with device-optimised and proven YOLO-driven object detection and segmentation techniques. Essentially, computational resource is focused onto spatio-temporal regions of high uncertainty to significantly improve UAV processing speeds. Alongside, we introduce our WildLive dataset, which comprises 200K+ annotated animal instances across 19K+ frames from 4K UAV videos collected at the Ol Pejeta Conservancy in Kenya. All frames contain ground truth bounding boxes, segmentation masks, as well as individual tracklets and tracking point trajectories. We compare our system against current object tracking approaches including OC-SORT, ByteTrack, and SORT. Our multi-animal tracking experiments with onboard hardware confirm that near real-time high-resolution wildlife tracking is possible on UAVs whilst maintaining high accuracy levels as needed for future navigational and mission-specific animal-centric operational autonomy. Our materials are available at: https://dat-nguyenvn.github.io/WildLive/

cs.CV↗

arXiv2025

MMLA: Multi-Environment, Multi-Species, Low-Altitude Drone Dataset

Real-time wildlife detection in drone imagery supports critical ecological and conservation monitoring. However, standard detection models like YOLO often fail to generalize across locations and struggle with rare species, limiting their use in automated drone deployments. We present MMLA, a novel multi-environment, multi-species, low-altitude drone dataset collected across three sites (Ol Pejeta Conservancy and Mpala Research Centre in Kenya, and The Wilds in Ohio), featuring six species (zebras, giraffes, onagers, and African wild dogs). The dataset contains 811K annotations from 37 high-resolution videos. Baseline YOLO models show performance disparities across locations while fine-tuning YOLOv11m on MMLA improves mAP50 to 82%, a 52-point gain over baseline. Our results underscore the need for diverse training data to enable robust animal detection in autonomous drone systems.

cs.CV↗

arXiv2024

KerasCV and KerasNLP: Vision and Language Power-Ups

We present the Keras domain packages KerasCV and KerasNLP, extensions of the Keras API for Computer Vision and Natural Language Processing workflows, capable of running on either JAX, TensorFlow, or PyTorch. These domain packages are designed to enable fast experimentation, with a focus on ease-of-use and performance. We adopt a modular, layered design: at the library's lowest level of abstraction, we provide building blocks for creating models and data preprocessing pipelines, and at the library's highest level of abstraction, we provide pretrained ``task" models for popular architectures such as Stable Diffusion, YOLOv8, GPT2, BERT, Mistral, CLIP, Gemma, T5, etc. Task models have built-in preprocessing, pretrained weights, and can be fine-tuned on raw inputs. To enable efficient training, we support XLA compilation for all models, and run all preprocessing via a compiled graph of TensorFlow operations using the tf.data API. The libraries are fully open-source (Apache 2.0 license) and available on GitHub.

cs.AI↗

arXiv2024

AttenCraft: Attention-guided Disentanglement of Multiple Concepts for Text-to-Image Customization

Text-to-image (T2I) customization empowers users to adapt the T2I diffusion model to new concepts absent in the pre-training dataset. On this basis, capturing multiple new concepts from a single image has emerged as a new task, allowing the model to learn multiple concepts simultaneously or discard unwanted concepts. However, multiple-concept disentanglement remains a key challenge. Existing disentanglement models often exhibit two main issues: feature fusion and asynchronous learning across different concepts. To address these issues, we propose AttenCraft, an attention-based method for multiple-concept disentanglement. Our method uses attention maps to generate accurate masks for each concept in a single initialization step, aiding in concept disentanglement without requiring mask preparation from humans or specialized models. Moreover, we introduce an adaptive algorithm based on attention scores to estimate sampling ratios for different concepts, promoting balanced feature acquisition and synchronized learning. AttenCraft also introduces a feature-retaining training framework that employs various loss functions to enhance feature recognition and prevent fusion. Extensive experiments show that our model effectively mitigates these two issues, achieving state-of-the-art image fidelity and comparable prompt fidelity to baseline models.

cs.CV↗

arXiv2024

Grand Design vs. Multi-Armed Spiral Galaxies: Dependence on Galaxy Structure

We developed an algorithm to use Galaxy Zoo 3D spiral arm masks produced by citizen scientist volunteers to semi-automatically classify spiral galaxies as either multi-armed or grand design spirals. Our final sample consists of 299 multi-armed and 245 grand design galaxies. On average, the grand design galaxies have smaller stellar masses than the multi-armed galaxies. For a given stellar mass, the grand design galaxies have larger concentrations, earlier Hubble types, smaller half-light radii, and larger central surface mass densities than the multi-armed galaxies. Lower mass galaxies of both arm classes have later Hubble types and lower concentrations than higher mass galaxies. In our sample, a higher fraction of grand design galaxies have classical bulges rather than pseudo-bulges, compared to multi-armed galaxies. These results are consistent with theoretical models and simulations which suggest that dense classical bulges support the development and/or longevity of 2-armed spiral patterns. Similar specific star formation rates are found in multi-armed and grand design galaxies with similar stellar masses and concentrations. This implies that the specific star formation rates in spiral galaxies is a function of concentration and stellar mass, but independent of the number of spiral arms. Our classifications are consistent with arm counts from the Galaxy Zoo 2 project and published m=3 Fourier amplitudes.

astro-ph.GA↗

arXiv2024

Textual Localization: Decomposing Multi-concept Images for Subject-Driven Text-to-Image Generation

Subject-driven text-to-image diffusion models empower users to tailor the model to new concepts absent in the pre-training dataset using a few sample images. However, prevalent subject-driven models primarily rely on single-concept input images, facing challenges in specifying the target concept when dealing with multi-concept input images. To this end, we introduce a textual localized text-to-image model (Texual Localization) to handle multi-concept input images. During fine-tuning, our method incorporates a novel cross-attention guidance to decompose multiple concepts, establishing distinct connections between the visual representation of the target concept and the identifier token in the text prompt. Experimental results reveal that our method outperforms or performs comparably to the baseline models in terms of image fidelity and image-text alignment on multi-concept input images. In comparison to Custom Diffusion, our method with hard guidance achieves CLIP-I scores that are 7.04%, 8.13% higher and CLIP-T scores that are 2.22%, 5.85% higher in single-concept and multi-concept generation, respectively. Notably, our method generates cross-attention maps consistent with the target concept in the generated images, a capability absent in existing models.

cs.CV↗

arXiv2022

One-dimensional electronic states in a natural misfit structure

Misfit compounds are thermodynamically stable stacks of two-dimensional materials, forming a three-dimensional structure that remains incommensurate in one direction parallel to the layers. As a consequence, no true bonding is expected between the layers, with their interaction being dominated by charge transfer. In contrast to this well-established picture, we show that interlayer coupling can strongly influence the electronic properties of one type of layer in a misfit structure, in a similar way to the creation of modified band structures in an artificial moir\'e structure between two-dimensional materials. Using angle-resolved photoemission spectroscopy with a micron-scale light focus, we selectively probe the electronic properties of hexagonal NbSe$_2$ and square BiSe layers that terminate the surface of the (BiSe)$_{1+\delta}$NbSe$_2$ misfit compound. We show that the band structure in the BiSe layers is strongly affected by the presence of the hexagonal NbSe$_2$ layers, leading to quasi one-dimensional electronic features. The electronic structure of the NbSe$_2$ layers, on the other hand, is hardly influenced by the presence of the BiSe. Using density functional theory calculations of the unfolded band structures, we argue that the preferred modification of one type of bands is mainly due to the atomic and orbital character of the states involved, opening a promising way to design novel electronic states that exploit the partially incommensurate character of the misfit compounds.

cond-mat.mtrl-sci↗

arXiv2021

Agree to Disagree: When Deep Learning Models With Identical Architectures Produce Distinct Explanations

Deep Learning of neural networks has progressively become more prominent in healthcare with models reaching, or even surpassing, expert accuracy levels. However, these success stories are tainted by concerning reports on the lack of model transparency and bias against some medical conditions or patients' sub-groups. Explainable methods are considered the gateway to alleviate many of these concerns. In this study we demonstrate that the generated explanations are volatile to changes in model training that are perpendicular to the classification task and model structure. This raises further questions about trust in deep learning models for healthcare. Mainly, whether the models capture underlying causal links in the data or just rely on spurious correlations that are made visible via explanation methods. We demonstrate that the output of explainability methods on deep neural networks can vary significantly by changes of hyper-parameters, such as the random seed or how the training set is shuffled. We introduce a measure of explanation consistency which we use to highlight the identified problems on the MIMIC-CXR dataset. We find explanations of identical models but with different training setups have a low consistency: $\approx$ 33% on average. On the contrary, kernel methods are robust against any orthogonal changes, with explanation consistency at 94%. We conclude that current trends in model explanation are not sufficient to mitigate the risks of deploying models in real life healthcare applications.

cs.LG↗

arXiv2021

Attack-agnostic Adversarial Detection on Medical Data Using Explainable Machine Learning

Explainable machine learning has become increasingly prevalent, especially in healthcare where explainable models are vital for ethical and trusted automated decision making. Work on the susceptibility of deep learning models to adversarial attacks has shown the ease of designing samples to mislead a model into making incorrect predictions. In this work, we propose a model agnostic explainability-based method for the accurate detection of adversarial samples on two datasets with different complexity and properties: Electronic Health Record (EHR) and chest X-ray (CXR) data. On the MIMIC-III and Henan-Renmin EHR datasets, we report a detection accuracy of 77% against the Longitudinal Adversarial Attack. On the MIMIC-CXR dataset, we achieve an accuracy of 88%; significantly improving on the state of the art of adversarial detection in both datasets by over 10% in all settings. We propose an anomaly detection based method using explainability techniques to detect adversarial samples which is able to generalise to different attack methods without a need for retraining.

cs.LG↗