arXiv subjects

Anthony Scanlan

Publications and source records attributed to Anthony Scanlan.

9 recordsLinked to original sources

arXiv2026

Probing Visual Concepts in Lightweight Vision-Language Models for Automated Driving

The use of Vision-Language Models (VLMs) in automated driving applications is becoming increasingly common, with the aim of leveraging their reasoning and generalisation capabilities to handle long-tail scenarios. However, these models often fail on simple visual questions that are highly relevant to automated driving, and the reasons behind these failures remain poorly understood. In this work, we examine the intermediate activations of VLMs and assess the extent to which specific visual concepts are linearly encoded, with the goal of identifying bottlenecks in the flow of visual information. Specifically, we create counterfactual image sets that differ only in a targeted visual concept and then train linear probes to distinguish between them using the activations of five state-of-the-art (SOTA) VLMs, including two training variants for one model. Our results show that concepts such as the presence of an object or agent in a scene are explicitly and linearly encoded, whereas other spatial visual concepts, such as the orientation of an object or agent, are only implicitly encoded by the spatial structure retained by the vision encoder, or are not linearly encoded at all. In parallel, we observe that in certain cases, even when a concept is linearly encoded in the model's activations, the model still fails to answer correctly. This leads us to identify two failure modes. The first is perceptual failure, where the visual information required to answer a question is not linearly encoded in the model's activations. The second is cognitive failure, where the visual information is present but the model fails to align it correctly with language semantics. Finally, we show that increasing the distance of the object in question quickly degrades the linear separability of the corresponding visual concept...

cs.CV↗

arXiv2025

Descriptor: Distance-Annotated Traffic Perception Question Answering (DTPQA)

The remarkable progress of Vision-Language Models (VLMs) on a variety of tasks has raised interest in their application to automated driving. However, for these models to be trusted in such a safety-critical domain, they must first possess robust perception capabilities, i.e., they must be capable of understanding a traffic scene, which can often be highly complex, with many things happening simultaneously. Moreover, since critical objects and agents in traffic scenes are often at long distances, we require systems with not only strong perception capabilities at close distances (up to 20 meters), but also at long (30+ meters) range. Therefore, it is important to evaluate the perception capabilities of these models in isolation from other skills like reasoning or advanced world knowledge. Distance-Annotated Traffic Perception Question Answering (DTPQA) is a Visual Question Answering (VQA) benchmark designed specifically for this purpose: it can be used to evaluate the perception systems of VLMs in traffic scenarios using trivial yet crucial questions relevant to driving decisions. It consists of two parts: a synthetic benchmark (DTP-Synthetic) created using a simulator, and a real-world benchmark (DTP-Real) built on top of existing images of real traffic scenes. Additionally, DTPQA includes distance annotations, i.e., how far the object in question is from the camera. More specifically, each DTPQA sample consists of (at least): (a) an image, (b) a question, (c) the ground truth answer, and (d) the distance of the object in question, enabling analysis of how VLM performance degrades with increasing object distance. In this article, we provide the dataset itself along with the Python scripts used to create it, which can be used to generate additional data of the same kind.

cs.CV↗

arXiv2025

Evaluating Small Vision-Language Models on Distance-Dependent Traffic Perception

Vision-Language Models (VLMs) are becoming increasingly powerful, demonstrating strong performance on a variety of tasks that require both visual and textual understanding. Their strong generalisation abilities make them a promising component for automated driving systems, which must handle unexpected corner cases. However, to be trusted in such safety-critical applications, a model must first possess a reliable perception system. Moreover, since critical objects and agents in traffic scenes are often at a distance, we require systems that are not "shortsighted", i.e., systems with strong perception capabilities at both close (up to 20 meters) and long (30+ meters) range. With this in mind, we introduce Distance-Annotated Traffic Perception Question Answering (DTPQA), the first Visual Question Answering (VQA) benchmark focused solely on perception-based questions in traffic scenes, enriched with distance annotations. By excluding questions that require reasoning, we ensure that model performance reflects perception capabilities alone. Since automated driving hardware has limited processing power and cannot support large VLMs, our study centers on smaller VLMs. More specifically, we evaluate several state-of-the-art (SOTA) small VLMs on DTPQA and show that, despite the simplicity of the questions, these models significantly underperform compared to humans (~60% average accuracy for the best-performing small VLM versus ~85% human performance). However, it is important to note that the human sample size was relatively small, which imposes statistical limitations. We also identify specific perception tasks, such as distinguishing left from right, that remain particularly challenging for these models.

cs.CV↗

arXiv2025

FIN: Fast Inference Network for Map Segmentation

Multi-sensor fusion in autonomous vehicles is becoming more common to offer a more robust alternative for several perception tasks. This need arises from the unique contribution of each sensor in collecting data: camera-radar fusion offers a cost-effective solution by combining rich semantic information from cameras with accurate distance measurements from radar, without incurring excessive financial costs or overwhelming data processing requirements. Map segmentation is a critical task for enabling effective vehicle behaviour in its environment, yet it continues to face significant challenges in achieving high accuracy and meeting real-time performance requirements. Therefore, this work presents a novel and efficient map segmentation architecture, using cameras and radars, in the \acrfull{bev} space. Our model introduces a real-time map segmentation architecture considering aspects such as high accuracy, per-class balancing, and inference time. To accomplish this, we use an advanced loss set together with a new lightweight head to improve the perception results. Our results show that, with these modifications, our approach achieves results comparable to large models, reaching 53.5 mIoU, while also setting a new benchmark for inference time, improving it by 260\% over the strongest baseline models.

cs.CV↗

arXiv2025

PAN: Pillars-Attention-Based Network for 3D Object Detection

Camera-radar fusion offers a robust and low-cost alternative to Camera-lidar fusion for the 3D object detection task in real-time under adverse weather and lighting conditions. However, currently, in the literature, it is possible to find few works focusing on this modality and, most importantly, developing new architectures to explore the advantages of the radar point cloud, such as accurate distance estimation and speed information. Therefore, this work presents a novel and efficient 3D object detection algorithm using cameras and radars in the bird's-eye-view (BEV). Our algorithm exploits the advantages of radar before fusing the features into a detection head. A new backbone is introduced, which maps the radar pillar features into an embedded dimension. A self-attention mechanism allows the backbone to model the dependencies between the radar points. We are using a simplified convolutional layer to replace the FPN-based convolutional layers used in the PointPillars-based architectures with the main goal of reducing inference time. Our results show that with this modification, our approach achieves the new state-of-the-art in the 3D object detection problem, reaching 58.2 of the NDS metric for the use of ResNet-50, while also setting a new benchmark for inference time on the nuScenes dataset for the same category.

cs.CV↗

arXiv2025

SOLAS: Superpositioning an Optical Lens in Automotive Simulation

Automotive Simulation is a potentially cost-effective strategy to identify and test corner case scenarios in automotive perception. Recent work has shown a significant shift in creating realistic synthetic data for road traffic scenarios using a video graphics engine. However, a gap exists in modeling realistic optical aberrations associated with cameras in automotive simulation. This paper builds on the concept from existing literature to model optical degradations in simulated environments using the Python-based ray-tracing library KrakenOS. As a novel pipeline, we degrade automotive fisheye simulation using an optical doublet with +/-2 deg Field of View (FOV), introducing realistic optical artifacts into two simulation images from SynWoodscape and Parallel Domain Woodscape. We evaluate KrakenOS by calculating the Root Mean Square Error (RMSE), which averaged around 0.023 across the RGB light spectrum compared to Ansys Zemax OpticStudio, an industrial benchmark for optical design and simulation. Lastly, we measure the image sharpness of the degraded simulation using the ISO12233:2023 Slanted Edge Method and show how both qualitative and measured results indicate the extent of the spatial variation in image sharpness from the periphery to the center of the degradations.

eess.SP↗

arXiv2024

Deformable Convolution Based Road Scene Semantic Segmentation of Fisheye Images in Autonomous Driving

This study investigates the effectiveness of modern Deformable Convolutional Neural Networks (DCNNs) for semantic segmentation tasks, particularly in autonomous driving scenarios with fisheye images. These images, providing a wide field of view, pose unique challenges for extracting spatial and geometric information due to dynamic changes in object attributes. Our experiments focus on segmenting the WoodScape fisheye image dataset into ten distinct classes, assessing the Deformable Networks' ability to capture intricate spatial relationships and improve segmentation accuracy. Additionally, we explore different loss functions to address class imbalance issues and compare the performance of conventional CNN architectures with Deformable Convolution-based CNNs, including Vanilla U-Net and Residual U-Net architectures. The significant improvement in mIoU score resulting from integrating Deformable CNNs demonstrates their effectiveness in handling the geometric distortions present in fisheye imagery, exceeding the performance of traditional CNN architectures. This underscores the significant role of Deformable convolution in enhancing semantic segmentation performance for fisheye imagery.

cs.CV↗

arXiv2024

Surround-View Fisheye Optics in Computer Vision and Simulation: Survey and Challenges

In this paper, we provide a survey on automotive surround-view fisheye optics, with an emphasis on the impact of optical artifacts on computer vision tasks in autonomous driving and ADAS. The automotive industry has advanced in applying state-of-the-art computer vision to enhance road safety and provide automated driving functionality. When using camera systems on vehicles, there is a particular need for a wide field of view to capture the entire vehicle's surroundings, in areas such as low-speed maneuvering, automated parking, and cocoon sensing. However, one crucial challenge in surround-view cameras is the strong optical aberrations of the fisheye camera, which is an area that has received little attention in the literature. Additionally, a comprehensive dataset is needed for testing safety-critical scenarios in vehicle automation. The industry has turned to simulation as a cost-effective strategy for creating synthetic datasets with surround-view camera imagery. We examine different simulation methods (such as model-driven and data-driven simulations) and discuss the simulators' ability (or lack thereof) to model real-world optical performance. Overall, this paper highlights the optical aberrations in automotive fisheye datasets, and the limitations of optical reality in simulated fisheye datasets, with a focus on computer vision in surround-view optical systems.

cs.CV↗

arXiv2024

Measuring Natural Scenes SFR of Automotive Fisheye Cameras

The Modulation Transfer Function (MTF) is an important image quality metric typically used in the automotive domain. However, despite the fact that optical quality has an impact on the performance of computer vision in vehicle automation, for many public datasets, this metric is unknown. Additionally, wide field-of-view (FOV) cameras have become increasingly popular, particularly for low-speed vehicle automation applications. To investigate image quality in datasets, this paper proposes an adaptation of the Natural Scenes Spatial Frequency Response (NS-SFR) algorithm to suit cameras with a wide field-of-view.

cs.CV↗