Effectively capturing intricate interactions among road users plays a critical role in achieving safe navigation for autonomous vehicles. While graph learning (GL) has emerged as a promising approach to tackle this challenge, existing GL models rely on predefined domain-specific graph extraction rules and often fail in real-world dynamic scenarios. Additionally, these graph extraction rules severely impede the capability of existing GL methods to generalize knowledge across domains. To address this issue, we propose RoadScene2Graph (RS2G), an innovative autonomous scenario understanding framework with a novel data-driven graph extraction and modeling approach that dynamically captures the diverse relations among road users. Our evaluations show that on average RS2G outperforms the state-of-the-art (SOTA) rule-based graph extraction method by 4.47% and the SOTA deep learning model by 22.19% in subjective risk assessment. RS2G also delivers notably better performance in transferring knowledge gained from simulations to unseen real-world scenarios.

Autonomous vehicles are cyber-physical systems that combine embedded computing and deep learning with physical systems to perceive the world, predict future states, and safely control the vehicle through changing environments. The ability of an autonomous vehicle to accurately predict the motion of other road users across a wide range of diverse scenarios is critical for both motion planning and safety. However, existing motion prediction methods do not explicitly model contextual information about the environment, which can cause significant variations in performance across diverse driving scenarios. To address this limitation, we propose CASTNet : a dynamic, context-aware approach for motion prediction that (i) identifies the current driving context using a spatio-temporal model, (ii) adapts an ensemble of motion prediction models to fit the current context, and (iii) applies novel trajectory fusion methods to combine predictions output by the ensemble. This approach enables CASTNet to improve robustness by minimizing motion prediction error across diverse driving scenarios. CASTNet is highly modular and can be used with various existing image processing backbones and motion predictors. We demonstrate how CASTNet can improve both CNN-based and graph-learning-based motion prediction approaches and conduct ablation studies on the performance, latency, and model size for various ensemble architecture choices. In addition, we propose and evaluate several attention-based spatio-temporal models for context identification and ensemble selection. We also propose a modular trajectory fusion algorithm that effectively filters, clusters, and fuses the predicted trajectories output by the ensemble. On the nuScenes dataset, our approach demonstrates more robust and consistent performance across diverse, real-world driving contexts than state-of-the-art techniques.

Autonomous systems (AS) are systems that can adapt and change their behaviors in response to unanticipated events and include systems such as aerial drones, autonomous vehicles, and ground/aquatic robots. AS require a wide array of sensors, deep learning models, and powerful hardware platforms to perceive the environment and safely operate in real-time. However, in many contexts, some sensing modalities negatively impact perception while increasing the system's overall energy consumption. Since AS are often energy-constrained edge devices, energy-efficient sensor fusion methods have been proposed. However, existing methods either fail to adapt to changing scenario conditions or to optimize system-wide energy efficiency. We propose CARMA, a context-aware sensor fusion approach that uses context to dynamically reconfigure the computation flow on a field-programmable gate array (FPGA) at runtime. By clock gating unused sensors and model sub-components, CARMA significantly reduces the energy used by a multi-sensory object detector without compromising performance. We use a deep learning processor unit (DPU) based reconfiguration approach to minimize the latency of model reconfiguration. We evaluate multiple context identification strategies, propose a novel system-wide energy-performance joint optimization, and evaluate scenario-specific perception performance. Across challenging real-world sensing contexts, CARMA outperforms state-of-the-art methods with up to 1.3× speedup and 73% lower energy consumption.

Autonomous vehicles use multiple sensors, large deep-learning models, and powerful hardware platforms to perceive the environment and navigate safely. In many contexts, some sensing modalities negatively impact perception while increasing energy consumption. We propose EcoFusion: an energy-aware sensor fusion approach that uses context to adapt the fusion method and reduce energy consumption without affecting perception performance. EcoFusion performs up to 9.5% better at object detection than existing fusion methods with approximately 60% less energy and 58% lower latency on the industry-standard Nvidia Drive PX2 hardware platform. We also propose several context-identification strategies, implement a joint optimization between energy and performance, and present scenario-specific results.

In autonomous vehicles (AVs), early warning systems rely on collision prediction to ensure occupant safety. However, state-of-the-art methods using deep convolutional networks either fail at modeling collisions or are too expensive/slow, making them less suitable for deployment on AV edge hardware. To address these limitations, we propose SG2VEC, a spatiotemporalscene-graphembedding methodology that uses the graph neural network (GNN) and long short-term memory (LSTM) layers to predict future collisions via visual scene perception. We demonstrate that SG2VEC predicts collisions 8.11% more accurately and 39.07% earlier than the state-of-the-art method on synthesized data sets, and 29.47% more accurately on a challenging real-world collision data set. We also show that SG2VEC is better than the state of the art at transferring knowledge from synthetic data sets to real-world driving data sets. Finally, we demonstrate that SG2VEC performs inference$9.3\times $faster with an 88.0% smaller model, 32.4% less power, and 92.8% less energy than the state-of-the-art method on the industry-standard Nvidia DRIVE PX 2 platform, making it more suitable for implementation on the edge.

Recently, road scene-graph representations used in conjunction with graph learning techniques have been shown to outperform state-of-the-art deep learning techniques in tasks including action classification, risk assessment, and collision prediction. To enable the exploration of applications of road scene-graph representations, we introduce roadscene2vec: an open-source tool for extracting and embedding road scene-graphs. The goal of roadscene2vec is to enable research into the applications and capabilities of road scene-graphs by providing tools for generating scene-graphs, graph learning models to create spatio-temporal scene-graph embeddings, and tools for visualizing and analyzing scene-graph-based methodologies. The capabilities of roadscene2vec include (i) customized scene-graph generation from either video clips or data from the CARLA simulator, (ii) multiple configurable spatio-temporal graph embedding models and baseline CNN-based models, (iii) built-in functionality for using graph and sequence embeddings for risk assessment and collision prediction applications, (iv) tools for evaluating transfer learning, and (v) utilities for visualizing scene-graphs and analyzing the explainability of graph learning models. We demonstrate the utility of roadscene2vec for these use cases with experimental results and qualitative evaluations for both graph learning models and CNN-based models. roadscene2vec is available at https://github.com/AICPS/roadscene2vec.

There is considerable evidence that evaluating the subjective risk level of driving decisions can improve the safety of Autonomous Driving Systems (ADS) in both typical and complex driving scenarios. In this paper, we propose a novel data-driven approach that uses scene-graphs as intermediate representations for modeling the subjective risk of driving maneuvers. Our approach includes a Multi-Relation Graph Convolution Network, a Long-Short Term Memory Network, and attention layers. To train our model, we formulate subjective risk assessment as a supervised scene classification problem. We evaluate our model on both synthetic lane-changing datasets and real-driving datasets with various driving maneuvers. We show that our approach achieves a higher classification accuracy than the state-of-the-art approach on both large (96.4% vs. 91.2%) and small (91.8% vs. 71.2%) lane-changing synthesized datasets, illustrating that our approach can learn effectively even from small datasets. We also show that our model trained on a lane-changing synthesized dataset achieves an average accuracy of 87.8% when tested on a real-driving lane-changing dataset. In comparison, the state-of-the-art model trained on the same synthesized dataset only achieved 70.3% accuracy when tested on the real-driving dataset, showing that our approach can transfer knowledge more effectively. Moreover, we demonstrate that the addition of spatial and temporal attention layers improves our model’s performance and explainability. Finally, our results illustrate that our model can assess the risk of various driving maneuvers more accurately than the state-of-the-art model (86.5% vs. 58.4%, respectively).

Autonomous vehicles (AV) are expected to revolutionize transportation and improve road safety significantly. However, these benefits do not come without cost; AVs require large Deep-Learning (DL) models and powerful hardware platforms to operate reliably in real-time, requiring between several hundred watts to one kilowatt of power. This power consumption can dramatically reduce vehicles’ driving range and affect emissions. To address this problem, we propose SAGE: a methodology for selectively offloading the key energy-consuming modules of DL architectures to the cloud to optimize edge, energy usage while meeting real-time latency constraints. Furthermore, we leverage Head Network Distillation (HND) to introduce efficient bottlenecks within the DL architecture in order to minimize the network overhead costs of offloading with almost no degradation in the model’s performance. We evaluate SAGE using an Nvidia Jetson TX2 and an industry-standard Nvidia Drive PX2 as the AV edge, devices and demonstrate that our offloading strategy is practical for a wide range of DL models and internet connection bandwidths on 3G, 4G LTE, and WiFi technologies. Compared to edge-only computation, SAGE reduces energy consumption by an average of 36.13% , 47.07% , and 55.66% for an AV with one low-resolution camera, one high-resolution camera, and three high-resolution cameras, respectively. SAGE also reduces upload data size by up to 98.40% compared to direct camera offloading.

If machine failures can be detected preemptively, then maintenance and repairs can be performed more efficiently, reducing production costs. Many machine learning techniques for performing early failure detection using vibration data have been proposed; however, these methods are often power and data-hungry, susceptible to noise, and require large amounts of data preprocessing. Also, training is usually only performed once before inference, so they do not learn and adapt as the machine ages. In this article, we propose a method of performing online, real-time anomaly detection for predictive maintenance using hierarchical temporal memory (HTM). Inspired by the human neocortex, HTMs learn and adapt continuously and are robust to noise. Using the Numenta Anomaly Benchmark, we empirically demonstrate that our approach outperforms state-of-the-art algorithms at preemptively detecting real-world cases of bearing failures and simulated 3-D printer failures. Our approach achieves an average score of 64.71, surpassing state-of-the-art deep-learning (49.38) and statistical (61.06) methods.

Due to the sensitivity of the information exchanged in Vehicle to Vehicle (V2V) and Vehicle to Infrastructure (V2I) communication, generating secret keys is critical to secure these communications. As nature is open access, distributed keys are more vulnerable to attacks in the vehicular environment. Physical layer key generation methods using wireless channel characteristics show promise in preventing such attacks, generating keys independently, and removing the need for distribution. In this paper, we present a novel key generation approach in a real vehicular environment based on Channel State Information (CSI), including a new algorithm for key bit extraction. We implemented our algorithm using USRP B210 Software-Defined Radios (SDR) and the industry-standard V2X communication protocol: IEEE 802.11p. The proposed key generation protocol uses the CSI values of each sub-carrier as a source of randomness, from which bits are extracted using a new QAM demodulator quantizer (QAM-Dem-Quan). We compared our technique to state-of-the-art Received Signal Strength (RSS)-based approaches, and show that our method achieves better performance. Moreover, we reached a min-entropy of approximately 70% for the generated keys and a key generation rate of less than$150~\mu \text{s}$/key for key lengths ranging from 16 to 128 bits.

Keeping up with the rapid growth of Deep Learning (DL) research is a daunting task. While existing scientific literature search systems provide text search capabilities and can identify similar papers, gaining an in-depth understanding of a new approach or an application is much more complicated. Many publications leverage multiple modalities to convey their findings and spread their ideas - they include pseudocode, tables, images and diagrams in addition to text, and often make publicly accessible their implementations. It is important to be able to represent and query them as well. We utilize RDF Knowledge graphs (KGs) to represent multimodal information and enable expressive querying over modalities. In our demo we present an approach for extracting KGs from different modalities, namely text, architecture images and source code. We show how graph queries can be used to get insights into different facets (modalities) of a paper, and its associated code implementation. Our innovation lies in the multimodal nature of the KG we create. While our work is of direct interest to DL researchers and practitioners, our approaches can also be leveraged in other scientific domains.