no code implementations • 28 Mar 2024 • Mingyu Cai, Karankumar Patel, Soshi Iba, Songpo Li
We implement the algorithm on a virtual reality (VR) setup to teleoperate robotic hands in a simulation with various assembly tasks to show the effectiveness of online estimation.
no code implementations • 15 Sep 2023 • Kaier Liang, Mingyu Cai, Cristian-Ioan Vasile
We developed an explicit reference governor-guided control barrier function (ERG-guided CBF) method that enables the application of first-order CBFs to high-order linearizable systems.
1 code implementation • 30 Apr 2023 • Junchao Li, Mingyu Cai, Zhen Kan, Shaoping Xiao
We formulate motion planning as a probabilistic-labeled partially observable Markov decision process (PL-POMDP) problem and use linear temporal logic (LTL) to express the complex task.
1 code implementation • 4 Oct 2022 • Danyang Li, Mingyu Cai, Cristian-Ioan Vasile, Roberto Tron
Machine learning techniques using neural networks have achieved promising success for time-series data classification.
no code implementations • 3 Oct 2022 • Mingyu Cai, Makai Mann, Zachary Serlin, Kevin Leahy, Cristian-Ioan Vasile
This is achieved by decomposing an infeasible LTL formula into several reach-avoid sub-tasks with shorter horizons, which can be trained in a modular DRL architecture.
1 code implementation • 15 Sep 2022 • Zhangli Zhou, Shaochen Wang, Ziyang Chen, Mingyu Cai, Zhen Kan
We demonstrate that using parallel branches as opposed to serial stacked convolutional layers will be a more powerful design for robotic visual grasping tasks.
no code implementations • 28 Jan 2022 • Mingyu Cai, Erfan Aasi, Calin Belta, Cristian-Ioan Vasile
This work presents a deep policy gradient algorithm for controlling a robot with unknown dynamics operating in a cluttered environment when the task is specified as a Linear Temporal Logic (LTL) formula.
no code implementations • 28 Dec 2021 • Erfan Aasi, Mingyu Cai, Cristian Ioan Vasile, Calin Belta
In this paper, we introduce a time-incremental learning framework: given a dataset of labeled signal traces with a common time horizon, we propose a method to predict the label of a signal that is received incrementally over time, referred to as prefix signal.
1 code implementation • 27 Dec 2021 • Yue Zhu, Mingyu Cai, Chris Schwarz, Junchao Li, Shaoping Xiao
At first, the obtained optimal policy from PPO is compared to those from DQN and DDQN.
no code implementations • 7 Sep 2021 • Mingyu Cai, Cristian-Ioan Vasile
Then, by applying a reward shaping technique, we develop a modular policy-gradient architecture exploiting the benefits of the automaton structure to decompose overall tasks and enhance the performance of learned controllers; (3) by incorporating Gaussian Processes (GPs) to estimate the uncertain dynamic systems, we synthesize a model-based safe exploration during the learning process using Exponential Control Barrier Functions (ECBFs) that generalize systems with high-order relative degrees; (4) to further improve the efficiency of exploration, we utilize the properties of LTL automata and ECBFs to propose a safe guiding process.
1 code implementation • 24 Feb 2021 • Mingyu Cai, Mohammadhosein Hasanbeig, Shaoping Xiao, Alessandro Abate, Zhen Kan
This paper investigates the motion planning of autonomous dynamical systems modeled by Markov decision processes (MDP) with unknown transition probabilities over continuous state and action spaces.
no code implementations • 25 Jan 2021 • Mingyu Cai, Shaoping Xiao, Zhijun Li, Zhen Kan
This paper studies the control synthesis of motion planning subject to uncertainties.
1 code implementation • 14 Oct 2020 • Mingyu Cai, Shaoping Xiao, Baoluo Li, Zhiliang Li, Zhen Kan
This paper presents a model-free reinforcement learning (RL) algorithm to synthesize a control policy that maximizes the satisfaction probability of linear temporal logic (LTL) specifications.