An official website of the United States government
3D LiDAR scanners are playing an increasingly important role in autonomous driving as they can generate depth information of the environment. However, creating large 3D LiDAR point cloud datasets with point-level labels requires a significant amount of manual annotation. This jeopardizes the efficient development of supervised deep learning algorithms which are often data-hungry. We present a framework to rapidly create point clouds with accurate pointlevel labels from a computer game. To our best knowledge, this is the first publication on LiDAR point cloud simulation framework for autonomous driving. The framework supports data collection from both auto-driving scenes and user-configured scenes. Point clouds from auto-driving scenes can be used as training data for deep learning algorithms, while point clouds from user-configured scenes can be used to systematically test the vulnerability of a neural network, and use the falsifying examples to make the neural network more robust through retraining. In addition, the scene images can be captured simultaneously in order for sensor fusion tasks, with a method proposed to do automatic registration between the point clouds and captured scene images. We show a significant improvement in accuracy (+9%) in point cloud segmentation by augmenting the training dataset with the generated synthesized data. Our experiments also show by testing and retraining the network using point clouds from user-configured scenes, the weakness/blind spots of the neural network can be fixed.
more »
« less
ADAPTIVE ACQUISITION OF AIRBORNE LIDAR POINT CLOUD BASED ON DEEP REINFORCEMENT LEARNING
Human experience involvement in existing operations of airborne Light Detection and Ranging (LIDAR) systems and off-line processing of collected LIDAR data make the acquisition process of airborne LIDAR point cloud less adaptable to environment conditions. This work develops a deep reinforcement learning-enabled framework for adaptive airborne LIDAR point cloud acquisition. Namely, the optimization of the airborne LIDAR operation is modeled as a Markov decision process (MDP). A set of LIDAR point cloud processing methods are proposed to derive the state space, action space, and reward function of the MDP model. A DRL algorithm, Deep Q-Network (DQN), is used to solve the MDP. The DRL model is trained in a flexible virtual environment by using simulator AirSim. Extensive simulation demonstrates the efficiency of the proposed framework.
more »
« less
- Award ID(s):
- 1924278
- PAR ID:
- 10357308
- Date Published:
- Journal Name:
- 2022 IEEE 5th International Conference on Multimedia Information Processing and Retrieval (MIPR)
- Page Range / eLocation ID:
- 366 to 371
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
This paper presents a Multiplicative Extended Kalman Filter (MEKF) framework using a state-of-the-art velocimeter Light Detection and Ranging (LIDAR) sensor for Terrain Relative Navigation (TRN) applications. The newly developed velocimeter LIDAR is capable of providing simultaneous position, Doppler velocity, and reflectivity measurements for every point in the point cloud. This information, along with pseudo-measurements from point cloud registration techniques, a novel bulk velocity batch state estimation process and inertial measurement data, is fused within a traditional Kalman filter architecture. Results from extensive emulation robotics experiments performed at Texas A&M’s Land, Air, and Space Robotics (LASR) laboratory and Monte Carlo simulations are presented to evaluate the efficacy of the proposed algorithms.more » « less
-
Optimal sensor placement is critical for enhancing the effectiveness of monitoring dynamical systems. Deterministic solutions do not reflect the effects of input and parameter uncertainty on the sensor placement. Using a Markov decision process (MDP) and a sensor placement agent, this study proposes a stochastic approach to maximize the gain from placing a fixed number of sensors within the system. Utilizing Deep Reinforcement Learning (DRL), the agent is trained by collecting interactive samples within the environment, which uses an information-theoretic reward function that is a measure, based on Shannon entropy, of the identifiability of the model parameters. The goal of the agent is to maximize its expected future reward by selecting, at each step, the action (placing a sensor) that provides the most information. This framework is validated using a synthetic model of a base-isolated structure. To consider the existing uncertainty in the parameters, a prior probability distribution is chosen (e.g., based on expert judgement or preliminary study) for each model parameter. Further, a probabilistic model for the input is used to reflect input variability. In a Deep Q-network, a type of DRL algorithm, the agent learns a mapping from states (i.e., sensor configurations) to the "quality" of each action at that state, called "Q-values". This network is trained using samples of state, action, and reward by interacting with the environment. The modular property of the framework and the function approximation used in this study makes it scalable to complex real-world applications of sensor placement problems in the presence of uncertainties.more » « less
-
Abstract Supervised machine learning via artificial neural network (ANN) has gained significant popularity for many geomechanics applications that involves multi‐phase flow and poromechanics. For unsaturated poromechanics problems, the multi‐physics nature and the complexity of the hydraulic laws make it difficult to design the optimal setup, architecture, and hyper‐parameters of the deep neural networks. This paper presents a meta‐modeling approach that utilizes deep reinforcement learning (DRL) to automatically discover optimal neural network settings that maximize a pre‐defined performance metric for the machine learning constitutive laws. This meta‐modeling framework is cast as a Markov Decision Process (MDP) with well‐defined states (subsets of states representing the proposed neural network (NN) settings), actions, and rewards. Following the selection rules, the artificial intelligence (AI) agent, represented in DRL via NN, self‐learns from taking a sequence of actions and receiving feedback signals (rewards) within the selection environment. By utilizing the Monte Carlo Tree Search (MCTS) to update the policy/value networks, the AI agent replaces the human modeler to handle the otherwise time‐consuming trial‐and‐error process that leads to the optimized choices of setup from a high‐dimensional parametric space. This approach is applied to generate two key constitutive laws for the unsaturated poromechanics problems: (1) the path‐dependent retention curve with distinctive wetting and drying paths. (2) The flow in the micropores, governed by an anisotropic permeability tensor. Numerical experiments have shown that the resultant ML‐generated material models can be integrated into a finite element (FE) solver to solve initial‐boundary‐value problems as replacements of the hand‐craft constitutive laws.more » « less
-
Creating cave maps is an essential part of cave research. Traditional cartographic efforts are extremely time consuming and subjective, motivating the development of new techniques using terrestrial lidar scanners and mobile lidar systems. However, processing the large point clouds from these scanners to produce detailed, yet manageable “maps” remains a challenge. In this work, we present a methodology for synthesizing a basemap representing the cave floor from large scale point clouds, based on a case study of a SLAM-based lidar data acquisition from a cave system in the archaeological site of Las Cuevas, Belize. In 4 days of fieldwork, the 335 m length of the cave system was scanned, resulting in a point cloud of 4.1 billion points, with 1.6 billion points classified as part of the cave floor. This point cloud was processed to produce a basemap that can be used in GIS, where natural and anthropogenic features are clearly visible and can be traced to create accurate 2D maps similar to traditional cartography.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1109/MIPR54900.2022.00072
