Improved vehicle classification using ResNets and spatial pooling.
problem Fine-grained vehicle classification using ResNet architectures.
method Training ResNet-18, -34, and -50 on Comprehensive Cars dataset. Adding Spatially Weighted Pooling and localisation.
result Combining Spatially Weighted Pooling and localisation increases top-1 accuracy to 96.351%.
Improved particle filters enhance vehicle tracking accuracy.
problem Particle filters struggle with frequent, informative observations.
method Proposes particle filters that sample around recent observations.
result Significant improvement in accuracy and efficiency.
We solve 6-DoF localisation and 3D reconstruction using deep state-space models.
problem 6-DoF localisation and dense 3D reconstruction in spatial environments.
method Approximate Bayesian inference in a deep state-space model combining learning and domain knowledge.
result Near state-of-the-art performance on UAV flight data.
New algorithm localizes spoofing attackers using tropical geometry.
problem Localizing spoofing attackers in digital systems.
method Adaptive pruning algorithm based on tropical geometry.
result Adaptive pruning improves localisation accuracy.
Equivalent bicategories constructed from action Lie groupoids.
problem Equivalence of bicategories constructed from action Lie groupoids.
method Localizing at equivariant weak equivalences, surjective submersive equivariant weak equivalences, and all weak equivalences.
result Weak equivalences between action Lie groupoids are isomorphic to compositions of nice forms of equivariant weak equivalences.
The paper applies S1-localization to symplectic cohomology.
problem Equivariant symplectic cohomology relations.
method Localisation by pseudocycles and moduli space lifting.
result Relations between equivariant symplectic classes and Gromov-Witten invariants.
Convolutional neural network localizes OD and fovea in UWFoV-SLO images.
problem Localizing optic disc and fovea centers in ultra-widefield retinal images.
method Convolutional neural network trained on reflectance and autofluorescence images.
result 99.4% OD localisation accuracy and 99.1% fovea localisation accuracy.
Solves open problem on Lie groupoids equivalence.
problem Whether Lie groupoids Morita equivalent are diffeologically Morita equivalent.
method Localisation of 2-categories, anafunctors, Lie groupoids, diffeological groupoids.
result Two Lie groupoids diffeologically Morita equivalent are Morita equivalent in the Lie sense.
Geometric interpretations and localisation theory for Kane-Mele invariant.
problem Understanding the Kane-Mele invariant in three-dimensional fermionic systems.
method Homotopy theory, geometric interpretations, Mayer-Vietoris Theorem, bundle gerbes.
result Unified cohomological explanation for equivalence between discrete Pfaffian and local geometric computations.
The study explores stable diffeomorphism groups in 4-manifolds using localisation and invariants.
problem Understanding stable diffeomorphism groups in 4-manifolds.
method Localisation of n-manifolds, inverting connected sum construction, using Bauer--Furuta invariants.
result K3-stable Bauer--Furuta invariants determine S^2xS^2-stable invariants.
Develops a new theory of localization in algebraic geometry.
problem Understanding localizations in cohomological theories with open-closed structures.
method Categorical and algebro-geometric approach, focusing on torsors and refinements.
result Establishes compatibility with various algebraic operations and recovers classical results.
CNNs help diagnose diabetic retinopathy by localizing lesions.
problem Diabetic retinopathy diagnosis requires identifying lesions in fundus images.
method Post-attention technique (Grad-CAM) on deep learning models' penultimate layer.
result InceptionV3 model achieves best performance and localizes lesions better.
We propose a method that performs anomaly detection and localisation within heterogeneous data using a pairwise undirected mixed graphical model. The data are a mixture of categorical and quantitative variables, and the model is learned over a dataset that is supposed not to contain any anomaly. We then use the model o…
Formula for fixed points on noncompact spaces.
problem Calculating fixed points on noncompact manifolds.
method Equivariant index theorem, localised functional, asymptotically local operators.
result Obtained a new Lefschetz fixed-point formula.
The paper calculates invariants for projective surfaces using Higgs pairs and virtual localisation.
problem Calculating invariants for projective surfaces with positive canonical bundle.
method Using Higgs pairs and virtual localisation, the paper defines invariants constant under deformations.
result The invariants can be rational and contribute to the Euler characteristic of the moduli space of instantons.
New approach to Z-stability and critical metrics on Kähler manifolds.
problem Determining Z-stability and existence of Z-critical metrics on Kähler manifolds. method Equivariant localisation applied to integrals over test configurations.
result Existence of Z-critical metrics is equivalent to Z-stability. Analytic torsion defined for non-compact Lie groups and discrete subgroups.
problem Defining and calculating analytic torsion for non-compact Lie groups and their discrete subgroups.
method Localised analytic torsion and relative analytic torsion defined for Lie groups of type I, using representations and discrete subgroups.
result Relative analytic torsion of (G,Γ) coincides with Lott L2 analytic torsion of a covering space. Deep learning tool classifies urban delivery vehicles.
problem Counting and categorizing delivery vehicles in cities.
method Developed annotated database and retrained CNNs.
result Accurate classification of 90%+ for 3 vehicle classes.
Paper tackles spooky effect in OSPA estimation, showing GOSPA solves it.
problem Spooky effect in optimal estimation of multiple targets with OSPA metric.
method Introduces GOSPA metric (α=2) to penalize false and missed targets. result GOSPA avoids the spooky effect and optimally lowers target estimation errors.
Enhances KWS in vehicles with multi-source fusion.
problem Improving precision and recall rates in vehicle keyword spotting.
method Integrates vehicle information into a DNN for speech classification and selects optimal sensitivity parameters.
result Significantly improved performance metrics (precision, recall, MSE) compared to baseline.
Adaptive framework generates challenging adversarial scenarios for autonomous vehicles.
problem Lack of efficient and adaptable evaluation methods for autonomous vehicles.
method Adaptive evaluation framework using ensemble models and nonparametric Bayesian clustering.
result Adversarial scenarios significantly degrade tested autonomous vehicles' performance.
Self-driving vehicles improve safety by predicting surrounding vehicles' trajectories.
problem Ensuring safety of self-driving vehicles through better trajectory prediction.
method Developed a Convolutional Neural Network to forecast vehicle trajectories from raw data.
result Improvement over baseline models in trajectory forecasting accuracy.
VTrackIt creates a synthetic dataset with infrastructure and vehicle info for AVs.
problem Lack of infrastructure and pooled vehicle info in existing AV datasets.
method Developed VTrackIt, a synthetic dataset with intelligent infrastructure and pooled vehicle info, and introduced InfraGAN for trajectory predictions.
result VTrackIt reduces high-risk edge cases in AV trajectory predictions.
Extends coarse index theory to locally compact groups for Callias operators.
problem Developing an equivariant coarse index theory for non-cocompact actions.
method Using admissible modules and localised K-theory of group C∗-algebras. result Equivariant index for Callias operators is a special case of the localised index.
New ensemble methods improve time series forecasting accuracy.
problem Global Forecasting Models (GFM) lack localisation for heterogeneous datasets.
method Ensemble techniques with clustering and varied GFM models.
result Significantly higher accuracy achieved compared to baseline models.
Proposes a privacy-preserving system for federated learning of road networks.
problem Privacy and security of data shared between vehicles and infrastructure.
method Federated learning over V2V and V2N links, non-IID dataset modeling.
result Improves learning performance and prevents eavesdropping.
Meta-models predict model hyperparameters for NDT experiments.
problem Non-destructive testing experiments are isolated; this work connects them.
method Bayesian multilevel approach, capturing inter-task relationships.
result Meta-models encode knowledge between and within tasks for transfer learning.
Generative model predicts vehicle faults up to 1000 hours in advance.
problem Forecasting vehicle faults for predictive maintenance.
method Generative model trained on US Army data, incorporating real-world factors.
result Highly accurate predictions of time to first fault.
Deep learning models control vehicle dynamics on a track.
problem Coupled longitudinal and lateral control of a vehicle.
method Trained two neural networks (MLP and CNN) to compute controls based on high-fidelity simulations.
result Deep learning models outperform conventional controllers on a challenging track.
This paper explores estimating chaotic dynamics and parameters using local ensemble Kalman filters.
problem Estimating chaotic dynamics and parameters from observations.
method Local ensemble Kalman filters with covariance and local domain localisation.
result Rigorously updating global parameters using a local domain ensemble Kalman filter.
Study SL(2,C) connections on Seifert-fibered spaces using gauge theory.
problem Counting SL(2,C) connections on Seifert-fibered spaces. method Introduced perturbations of the SL(2,C) Chern--Simons functional and proved a localisation result. result Formulae for the Euler characteristic and Poincaré polynomial of the stable locus of the SL(2,C) character variety of a Seifert-fibered homology 3-sphere. Generative model learns vehicle trajectory distributions for better data generalization.
problem Data sparsity and privacy issues in urban vehicle trajectory analysis.
method Generative adversarial imitation learning framework for urban vehicle trajectory generation.
result TrajGAIL model produces synthetic trajectories similar to real ones, achieving significant performance gains.
Proposes a method to predict vehicle intentions and motion adaptively.
problem Accurately predicting vehicle behaviors in various traffic scenarios.
method Probabilistic framework based on deep neural network.
result Better long-term motion prediction performance.
Adaptive stress testing for autonomous vehicles identifies failure scenarios using reinforcement learning.
problem Identifying potential failure scenarios in autonomous vehicle decision-making systems.
method Formulated as a Markov decision process, used reinforcement learning (DRL) to find likely failure scenarios.
result Deep Reinforcement Learning (DRL) finds more likely failure scenarios with fewer simulator calls than Monte Carlo Tree Search (MCTS).
Proposes a method to model multi-vehicle interactions using Gaussian processes.
problem Challenges in modeling correlations between multiple road users over time.
method Uses a stochastic vector field model and non-parametric Bayesian learning.
result Captures motion patterns from complex multi-vehicle interactions without heroic prior assumptions.
Proposes CTSDG model for better vehicle intention prediction across domains.
problem Domain generalization for vehicle intention prediction in dynamic environments.
method Structural causal model with recurrent latent variable integration.
result Consistent improvement in prediction accuracy compared to state-of-the-art methods.
A method predicts driving intentions of human-driven vehicles for safer autonomous driving.
problem Predicting timely driving intentions of human-driven vehicles for autonomous vehicles in mixed traffic.
method A Hidden Markov Model (HMM) approach using continuous mobility features.
result HMMs trained with continuous mobility features improve prediction accuracy.
Automatically identifies vehicles from audio sensors without needing labeled data.
problem Vehicle recognition and classification from acoustic signals.
method Incremental reseeding of acoustic signatures using spectral embedding and clustering.
result Incremental reseeding accurately identifies individual vehicles from their acoustic signatures.
AVs learn from past experiences to improve future performance.
problem Challenging situations and unknown experiences for AVs.
method Transfer Learning and Organic Computing.
result Online Transfer Learning helps update knowledge as tasks evolve.
The paper reviews machine learning safety techniques for autonomous vehicles.
problem Challenges in machine learning safety for autonomous vehicles.
method Organizes practical safety techniques to complement engineering safety.
result Enhances dependability and safety of machine learning algorithms in autonomous vehicles.
Acoustic sensors identify vehicles using spectral embedding.
problem Vehicle recognition from roadside audio sensors.
method Extract frequency signatures, apply spectral embedding for dimensionality reduction.
result K-nearest neighbors achieve accurate vehicle identification after dimensionality reduction.
Defining Vafa-Witten invariants for semistable Higgs pairs on polarised surfaces.
problem Counting semistable Higgs pairs on projective surfaces.
method Virtual localisation applied to Mochizuki/Joyce-Song pairs, proving for deg KS<0. result Invariants for K3 surfaces calculated in terms of modular forms.
New approach uses dynamic programming to efficiently discover failures in autonomous vehicle simulations.
problem Efficiently discovering rare failure events in autonomous vehicle simulations.
method Approximate dynamic programming and scene decomposition to estimate failure distribution.
result Increased number of failures discovered compared to baseline approaches.
We present a simple one-parameter model for spatially localised evolving agents competing for spatially localised resources. The model considers selling agents able to evolve their pricing strategy in competition for a fixed market. Despite its simplicity, the model displays extraordinarily rich behavior. In addition t…
Paper presents an energy-efficient RL method for sensor networks.
problem Energy consumption in sensor networks for health monitoring.
method Adaptive Reinforcement Learning framework using SARSA algorithm.
result Achieves performance enhancement and energy savings over time.
Paper proposes a method to improve autonomous vehicle performance using synthetically generated images.
problem Limited access to real-world datasets for autonomous vehicle training in countries with scarce data.
method Synthetically generated images to augment and train neural networks on small datasets.
result About 10% improvement in model performance observed.
Paper automates car negotiation in intersections using Q-learning.
problem Automated vehicles negotiate with human-driven cars in intersections.
method Deep Q-learning applied to simulated traffic with various driver behaviors.
result 98% success rate in avoiding collisions with other vehicles.
Improved reinforcement learning for 3D games using SLAM and object detection.
problem Challenges in 3D game environments, especially partial observability and combinatorial spaces.
method Augmented Deep Q-Learning Network with SLAM and object detection for better policy learning.
result Our approach consistently learns better policies in 3D games like Doom.