Tan Ho

The vehicle is equipped with a 1080p camera, a Velodyne VLP-16 Lidar and an industrial PC with Intel i9-9900K CPU and NVIDIA RTX2070 Super GPU for collecting training data and running our method. The network takes as input the ego state information (steering angle feedback and wheel speed feedback), camera images and Lidar data, and output speed predictions and control commands to drive the vehicle. More details can be found in Chapter 3.6.4 in my thesis https://lbezone.hkust.edu.hk/bib/991013106451703412

As the first of its kind in the university, many are delighted to witness the rise of some new artificial intelligent (AI) talents in the HKUST-Kaisa Autonomous RC Car Racing Competition held in 1 April 2021. The competition was sponsored by Kaisa Group and co-organized by School of Engineering, Robotics Institute, Intelligent Autonomous Driving Center, HKUST-Kaisa Joint Institute and Entrepreneurship Center. I am the PI of this project.

Objective: Fibrin gel has been demonstrated as an effective model for primary hepatic stellate cell (HSC) culture. However, fibrin gel exhibits some limitations, such as low transparency and poor stability. In this study, we aimed to address these drawbacks by investigating the influence of CaCl2 and hyaluronic acid (HA) concentrations on fibrin gel transparency, stability, and its suitability for HSC culturing. Methods: Different concentrations of CaCl2 (0.01, 0.025, 0.05, and 0.075 mM) and HA (0, 0.5, and 1 mg/mL) were employed to compare the properties of formed fibrin gels, including transparency, gel formation capability, stability, and stiffness. Additionally, HSC culturing experiments were performed on the formed gels. Results: Fibrin gel supplemented with 0.025 mM CaCl2 and 0.5 mg/mL HA exhibited the best transparency and stability after 7 days compared to other concentrations. Upon retrieval from fibrin gel, 3T3 cells display division rates comparable to those cultured on a plastic surface, accompanied by a high viability percentage, as evaluated through trypan blue staining. Moreover, HSC culturing on the formed gel revealed that HA promoted HSC activation by reducing lipid droplets and enhancing the expression of collagen markers. Conclusion: The combination of 0.025 mM CaCl2 and 0.5 mg/mL HA is suitable for generating fibrin gel for cell culturing purposes.

Lipopolysaccharide (LPS) leakage from bacteria causes liver inflammation and directly activates hepatic stellate cells (HSCs), contributing to liver fibrosis development. The current study investigated the effect of an autophagy inhibitor (Chloroquine, CQ) on in vitro HSC activation induced by LPS. Primary HSCs were isolated and cultured from BALB/c mice. Cells were cultured for 24 hours in a medium containing LPS 200 ng/ml, CQ 5 μM, or LPS 200 ng/ml + CQ 5 μM. All treated cells were evaluated by immunocytochemistry for morphology, proliferation, expression of activation markers α-sma and collagen 1, storage of lipid droplets and autophagy markers lc3/p62 at days 1 (D1), 3 (D3) and 7 (D7) after treatment. Proliferation of primary HSCs and expression of activation markers (α-sma and collagen 1) caused by LPS induction were significantly inhibited by CQ treatment until D7 of culture. Additionally, CQ treatment preserves cell function in lipid droplet storage. This effect of CQ on LPS-induced HSC activation was consistent with the late-stage inhibition of autophagy demonstrated by the decrease in lc3 and increase in p62 at D1. CQ inhibits primary HSC activation in vitro when induced by LPS.

Traditional modular self-driving frameworks scale poorly in new scenarios, which usually require tedious hand-tuning of rules and parameters to maintain acceptable performance in all foreseeable occasions. Therefore, robust and safe self-driving using traditional frameworks is still challenging, especially in complex and dynamic environments. Recently, deep-learning based self-driving methods have shown promising results with better generalization capability but less hand engineering effort. However, most of the previous learning-based methods are trained and evaluated in limited driving scenarios with scattered tasks, such as lane-following, autonomous braking, and conditional driving. In this paper, we propose a graph-based deep network to achieve scalable self-driving that can handle massive traffic scenarios. Specifically, more than 7,000 km of evaluation is conducted in a high-fidelity driving simulator, in which our method can obey the traffic rules and safely navigate the vehicle in a large variety of urban, rural, and highway environments, including unprotected left turns, narrow roads, roundabouts, and pedestrian-rich intersections. The results also show that our method achieves better performance over the baselines in terms of success rate.

Traditional methods for autonomous driving are implemented with many building blocks from perception, planning and control, making them difficult to generalize to varied scenarios due to complex assumptions and interdependencies. Recently, the end-to-end driving method has emerged, which performs well and generalizes to new environments by directly learning from export-provided data. However, many existing methods on this topic neglect to check the confidence of the driving actions and the ability to recover from driving mistakes. In this paper, we develop an uncertainty-aware end-to-end trajectory generation method based on imitation learning. It can extract spatiotemporal features from the front-view camera images for scene understanding, and then generate collision-free trajectories several seconds into the future. The experimental results suggest that under various weather and lighting conditions, our network can reliably generate trajectories in different urban environments, such as turning at intersections and slowing down for collision avoidance. Furthermore, closed-loop driving tests suggest that the proposed method achieves better cross-scene/platform driving results than the state-of-the-art (SOTA) end-to-end control method, where our model can recover from off-center and off-orientation errors and capture 80% of dangerous cases with high uncertainty estimations.

All-day and all-weather navigation is a critical capability for autonomous driving, which requires proper reaction to varied environmental conditions and complex agent behaviors. Recently, with the rise of deep learning, end-to-end control for autonomous vehicles has been studied a lot. However, most works are solely based on visual information, which could be degraded by challenging illumination conditions, such as dim lights or total darkness. In addition, they usually generate and apply deterministic control commands without considering the uncertainties in the future. In this paper, based on imitation learning, we propose a probabilistic driving model with multi-perception capability utilizing the information from camera, lidar and radar. We further evaluate its driving performance on-line on our new driving benchmark, which includes various environments (urban and rural areas, weathers, time of day, etc) and dynamic obstacles (vehicles, pedestrians, motorcyclists, bicyclists, etc). The results suggest that our proposed model outperforms baselines and achieves excellent generalization performance in unseen environments with heavy traffic and extreme weather conditions.

Drifting is a complicated task for autonomous vehicle control. Most traditional methods in this area are based on motion equations derived by the understanding of vehicle dynamics, which is difficult to be modeled precisely. We propose a robust drift controller without explicit motion equations, which is based on the latest model-free deep reinforcement learning algorithm soft actor-critic. The drift control problem is formulated as a trajectory following task, where the error-based state and reward are designed. After being trained on tracks with different levels of difficulty, our controller is capable of making the vehicle drift through various sharp corners quickly and stably in the unseen map. The proposed controller is further shown to have excellent generalization ability, which can directly handle unseen vehicle types with different physical properties, such as mass, tire friction, etc.

Reliable trajectory planning like human drivers in real-world dynamic urban environments is a critical capability for autonomous driving. To this end, we develop a vision and imitation learning-based planner to generate collision-free trajectories several seconds into the future. Our network consists of three sub-networks to conduct three basic driving tasks: keep straight, turn left and turn right. During the planning process, high-level commands are received as prior information to select a specific sub-network. We create our dataset from the Robotcar dataset, and the experimental results suggest that our planner is able to reliably generate trajectories in various driving tasks, such as turning at different intersections, lane-keeping on curved roads and changing lanes for collision avoidance.