arXiv subjects

Sascha Hornauer

Publications and source records attributed to Sascha Hornauer.

14 recordsLinked to original sources

arXiv2026

Towards Zero-Shot Transfer Across Embodiments For Driving VLAs

Vision-Language-Action models (VLAs) have shown strong potential in autonomous driving by leveraging multimodal pretraining for instruction following, visual reasoning, and scene-level generalization. In robotic manipulation, scaling VLA fine-tuning across multiple robot setups--especially when unifying representations across embodiments--has been shown to improve in-dataset performance and cross-embodiment generalization; in autonomous driving, however, VLAs remain largely trained on individual datasets and are rarely evaluated for zero-shot transfer to unseen datasets and camera rigs; furthermore naively adding more datasets to the training data does not necessarily lead to better performance within seen embodiments. To address these problems, we study multi-dataset training for the driving task and BEV-Forcing, an auxiliary objective that transfers ground-plane object-layout information from a specialized Bird's-Eye-View model into the VLA backbone. By encouraging the model to represent object position through a shared BEV spatial interface, we show that an auxiliary task such as BEV-Forcing can improve both in-distribution and out-of-distribution performance when training on a small number of camera rigs. As the number of training embodiments increases, however, the benefits of the auxiliary task are reduced; we present this as evidence that new techniques in the literature may see their benefits diminish when simply scaling up training diversity, which motivates presenting results taking into account data scaling.

cs.CV↗

arXiv2026

Future-Interactions-Aware Trajectory Prediction via Braid Theory

To safely operate, an autonomous vehicle must know the future behavior of a potentially high number of interacting agents around it, a task often posed as multi-agent trajectory prediction. Many previous attempts to model social interactions and solve the joint prediction task either add extensive computational requirements or rely on heuristics to label multi-agent behavior types. Braid theory, in contrast, provides a powerful exact descriptor of multi-agent behavior by projecting future trajectories into braids that express how trajectories cross with each other over time; a braid then corresponds to a specific mode of coordination between the multiple agents in the future. In past work, braids have been used lightly to reason about interacting agents and restrict the attention window of predicted agents. We show that leveraging more fully the expressivity of the braid representation and using it to condition the trajectories themselves leads to even further gains in joint prediction performance, with negligible added complexity either in training or at inference time. We do so by proposing a novel auxiliary task, braid prediction, done in parallel with the trajectory prediction task. By classifying edges between agents into their correct crossing types in the braid representation, the braid prediction task is able to imbue the model with improved social awareness, which is reflected in joint predictions that more closely adhere to the actual multi-agent behavior. This simple auxiliary task allowed us to obtain significant improvements in joint metrics on three separate datasets. We show how the braid prediction task infuses the model with future intention awareness, leading to more accurate joint predictions. Code is available at github.com/caiocj1/traj-pred-braid-theory.

cs.AI↗

arXiv2025

NeRAF: 3D Scene Infused Neural Radiance and Acoustic Fields

Sound plays a major role in human perception. Along with vision, it provides essential information for understanding our surroundings. Despite advances in neural implicit representations, learning acoustics that align with visual scenes remains a challenge. We propose NeRAF, a method that jointly learns acoustic and radiance fields. NeRAF synthesizes both novel views and spatialized room impulse responses (RIR) at new positions by conditioning the acoustic field on 3D scene geometric and appearance priors from the radiance field. The generated RIR can be applied to auralize any audio signal. Each modality can be rendered independently and at spatially distinct positions, offering greater versatility. We demonstrate that NeRAF generates high-quality audio on SoundSpaces and RAF datasets, achieving significant performance improvements over prior methods while being more data-efficient. Additionally, NeRAF enhances novel view synthesis of complex scenes trained with sparse data through cross-modal learning. NeRAF is designed as a Nerfstudio module, providing convenient access to realistic audio-visual generation.

cs.SD↗

arXiv2025

Improving Consistency in Vehicle Trajectory Prediction Through Preference Optimization

Trajectory prediction is an essential step in the pipeline of an autonomous vehicle. Inaccurate or inconsistent predictions regarding the movement of agents in its surroundings lead to poorly planned maneuvers and potentially dangerous situations for the end-user. Current state-of-the-art deep-learning-based trajectory prediction models can achieve excellent accuracy on public datasets. However, when used in more complex, interactive scenarios, they often fail to capture important interdependencies between agents, leading to inconsistent predictions among agents in the traffic scene. Inspired by the efficacy of incorporating human preference into large language models, this work fine-tunes trajectory prediction models in multi-agent settings using preference optimization. By taking as input automatically calculated preference rankings among predicted futures in the fine-tuning process, our experiments--using state-of-the-art models on three separate datasets--show that we are able to significantly improve scene consistency while minimally sacrificing trajectory prediction accuracy and without adding any excess computational requirements at inference time.

cs.LG↗

arXiv2024

AVS-Net: Audio-Visual Scale Net for Self-supervised Monocular Metric Depth Estimation

Metric depth prediction from monocular videos suffers from bad generalization between datasets and requires supervised depth data for scale-correct training. Self-supervised training using multi-view reconstruction can benefit from large scale natural videos but not provide correct scale, limiting its benefits. Recently, reflecting audible Echoes off objects is investigated for improved depth prediction and was shown to be sufficient to reconstruct objects at scale even without a visual signal. Because Echoes travel at fixed speed, they have the potential to resolve ambiguities in object scale and appearance. However, predicting depth end-to-end from sound and vision cannot benefit from unsupervised depth prediction approaches, which can process large scale data without sound annotation. In this work we show how Echoes can benefit depth prediction in two ways: When learning metric depth learned from supervised data and as supervisory signal for scale-correct self-supervised training. We show how we can improve the predictions of several state-of-the-art approaches and how the method can scale-correct a self-supervised depth approach.

cs.CV↗

arXiv2024

Mesoscale Traffic Forecasting for Real-Time Bottleneck and Shockwave Prediction

Accurate real-time traffic state forecasting plays a pivotal role in traffic control research. In particular, the CIRCLES consortium project necessitates predictive techniques to mitigate the impact of data source delays. After the success of the MegaVanderTest experiment, this paper aims at overcoming the current system limitations and develop a more suited approach to improve the real-time traffic state estimation for the next iterations of the experiment. In this paper, we introduce the SA-LSTM, a deep forecasting method integrating Self-Attention (SA) on the spatial dimension with Long Short-Term Memory (LSTM) yielding state-of-the-art results in real-time mesoscale traffic forecasting. We extend this approach to multi-step forecasting with the n-step SA-LSTM, which outperforms traditional multi-step forecasting methods in the trade-off between short-term and long-term predictions, all while operating in real-time.

cs.LG↗

arXiv2024

The Audio-Visual BatVision Dataset for Research on Sight and Sound

Vision research showed remarkable success in understanding our world, propelled by datasets of images and videos. Sensor data from radar, LiDAR and cameras supports research in robotics and autonomous driving for at least a decade. However, while visual sensors may fail in some conditions, sound has recently shown potential to complement sensor data. Simulated room impulse responses (RIR) in 3D apartment-models became a benchmark dataset for the community, fostering a range of audiovisual research. In simulation, depth is predictable from sound, by learning bat-like perception with a neural network. Concurrently, the same was achieved in reality by using RGB-D images and echoes of chirping sounds. Biomimicking bat perception is an exciting new direction but needs dedicated datasets to explore the potential. Therefore, we collected the BatVision dataset to provide large-scale echoes in complex real-world scenes to the community. We equipped a robot with a speaker to emit chirps and a binaural microphone to record their echoes. Synchronized RGB-D images from the same perspective provide visual labels of traversed spaces. We sampled modern US office spaces to historic French university grounds, indoor and outdoor with large architectural variety. This dataset will allow research on robot echolocation, general audio-visual tasks and sound phænomena unavailable in simulated data. We show promising results for audio-only depth prediction and show how state-of-the-art work developed for simulated data can also succeed on our dataset. Project page: https://amandinebtto.github.io/Batvision-Dataset/

cs.RO↗

arXiv2023

MBAPPE: MCTS-Built-Around Prediction for Planning Explicitly

We present MBAPPE, a novel approach to motion planning for autonomous driving combining tree search with a partially-learned model of the environment. Leveraging the inherent explainable exploration and optimization capabilities of the Monte-Carlo Search Tree (MCTS), our method addresses complex decision-making in a dynamic environment. We propose a framework that combines MCTS with supervised learning, enabling the autonomous vehicle to effectively navigate through diverse scenarios. Experimental results demonstrate the effectiveness and adaptability of our approach, showcasing improved real-time decision-making and collision avoidance. This paper contributes to the field by providing a robust solution for motion planning in autonomous driving systems, enhancing their explainability and reliability.

cs.RO↗

arXiv2022

GRI: General Reinforced Imitation and its Application to Vision-Based Autonomous Driving

Deep reinforcement learning (DRL) has been demonstrated to be effective for several complex decision-making applications such as autonomous driving and robotics. However, DRL is notoriously limited by its high sample complexity and its lack of stability. Prior knowledge, e.g. as expert demonstrations, is often available but challenging to leverage to mitigate these issues. In this paper, we propose General Reinforced Imitation (GRI), a novel method which combines benefits from exploration and expert data and is straightforward to implement over any off-policy RL algorithm. We make one simplifying hypothesis: expert demonstrations can be seen as perfect data whose underlying policy gets a constant high reward. Based on this assumption, GRI introduces the notion of offline demonstration agents. This agent sends expert data which are processed both concurrently and indistinguishably with the experiences coming from the online RL exploration agent. We show that our approach enables major improvements on vision-based autonomous driving in urban environments. We further validate the GRI method on Mujoco continuous control tasks with different off-policy RL algorithms. Our method ranked first on the CARLA Leaderboard and outperforms World on Rails, the previous state-of-the-art, by 17%.

cs.RO↗

arXiv2021

Unsupervised Discriminative Learning of Sounds for Audio Event Classification

Recent progress in network-based audio event classification has shown the benefit of pre-training models on visual data such as ImageNet. While this process allows knowledge transfer across different domains, training a model on large-scale visual datasets is time consuming. On several audio event classification benchmarks, we show a fast and effective alternative that pre-trains the model unsupervised, only on audio data and yet delivers on-par performance with ImageNet pre-training. Furthermore, we show that our discriminative audio learning can be used to transfer knowledge across audio datasets and optionally include ImageNet pre-training.

cs.SD↗

arXiv2020

BatVision with GCC-PHAT Features for Better Sound to Vision Predictions

Inspired by sophisticated echolocation abilities found in nature, we train a generative adversarial network to predict plausible depth maps and grayscale layouts from sound. To achieve this, our sound-to-vision model processes binaural echo-returns from chirping sounds. We build upon previous work with BatVision that consists of a sound-to-vision model and a self-collected dataset using our mobile robot and low-cost hardware. We improve on the previous model by introducing several changes to the model, which leads to a better depth and grayscale estimation, and increased perceptual quality. Rather than using raw binaural waveforms as input, we generate generalized cross-correlation (GCC) features and use these as input instead. In addition, we change the model generator and base it on residual learning and use spectral normalization in the discriminator. We compare and present both quantitative and qualitative improvements over our previous BatVision model.

cs.CV↗

arXiv2020

BatVision: Learning to See 3D Spatial Layout with Two Ears

Many species have evolved advanced non-visual perception while artificial systems fall behind. Radar and ultrasound complement camera-based vision but they are often too costly and complex to set up for very limited information gain. In nature, sound is used effectively by bats, dolphins, whales, and humans for navigation and communication. However, it is unclear how to best harness sound for machine perception. Inspired by bats' echolocation mechanism, we design a low-cost BatVision system that is capable of seeing the 3D spatial layout of space ahead by just listening with two ears. Our system emits short chirps from a speaker and records returning echoes through microphones in an artificial human pinnae pair. During training, we additionally use a stereo camera to capture color images for calculating scene depths. We train a model to predict depth maps and even grayscale images from the sound alone. During testing, our trained BatVision provides surprisingly good predictions of 2D visual scenes from two 1D audio signals. Such a sound to vision system would benefit robot navigation and machine vision, especially in low-light or no-light conditions. Our code and data are publicly available.

cs.CV↗

arXiv2018

Learning to Roam Free from Small-Space Autonomous Driving with A Path Planner

Modern autonomous driving algorithms often rely on learning the mapping from visual inputs to steering actions from human driving data in a variety of scenarios and visual scenes. The required data collection is not only labor intensive, but such data are often noisy, inconsistent, and inflexible, as there is no differentiation between good and bad drivers, or between different driving intentions. We propose a new autonomous driving approach that learns roaming skills from an optimal path planner. Our model car practices reaching random target locations in a small room with obstacles, by following the optimal trajectory and executing the steering actions decided by a planner. We learn the associations of driving behaviours with depth images, instead of raw color images of the visual scene. This more universal spatial representation allows the learned driving skills to transfer immediately to novel environments with different visual appearances. Our model car trained in a simple room, void of many visual features, demonstrates surprisingly good driving performance in a cluttered office environment, avoiding collisions with novel obstacles and unseen layouts of drive-able space. Its performance on outdoor curbside driving is also on par with human driving.

cs.RO↗

arXiv2017

Fast Recurrent Fully Convolutional Networks for Direct Perception in Autonomous Driving

Deep convolutional neural networks (CNNs) have been shown to perform extremely well at a variety of tasks including subtasks of autonomous driving such as image segmentation and object classification. However, networks designed for these tasks typically require vast quantities of training data and long training periods to converge. We investigate the design rationale behind end-to-end driving network designs by proposing and comparing three small and computationally inexpensive deep end-to-end neural network models that generate driving control signals directly from input images. In contrast to prior work that segments the autonomous driving task, our models take on a novel approach to the autonomous driving problem by utilizing deep and thin Fully Convolutional Nets (FCNs) with recurrent neural nets and low parameter counts to tackle a complex end-to-end regression task predicting both steering and acceleration commands. In addition, we include layers optimized for classification to allow the networks to implicitly learn image semantics. We show that the resulting networks use 3x fewer parameters than the most recent comparable end-to-end driving network and 500x fewer parameters than the AlexNet variations and converge both faster and to lower losses while maintaining robustness against overfitting.

cs.CV↗