27 January 2020 in IEEE Transactions on Transportation Electrification
25 February 2026 in Nature Communications
Autonomous vehicles remain commercially limited largely due to safety performance stagnation. Existing deep learning, heavily reliant on failure data from rare safety-critical events, suffers from the seesaw effect—improvement in some scenarios causes regression in others. We introduce an innovative dense learning approach that prioritizes both informative failures and successes, informed by theoretical findings. Data is sampled proportionally to its contribution to the policy gradient and exposure frequency, excluding non-informative samples. This densifies the training dataset’s information, significantly reducing learning variance without bias, enabling tasks intractable for existing methods. To validate this, we trained a safety-critical driving agent for a highly automated vehicle using mixed reality on an urban test track. Results demonstrate that our approach breaks the performance stagnation, enhancing the model’s overall safety performance by one to two orders of magnitude. This marks a significant stride towards achieving human-level safety and widespread adoption for autonomous vehicles.
30 November 2025 in Transportation Research Part C: Emerging Technologies
13 April 2026 in IEEE Transactions on Pattern Analysis and Machine Intelligence
Motion prediction is essential for autonomous vehicles to navigate complex environments and anticipate the behavior of other traffic participants. As new driving scenarios emerge, models must be continually updated without retraining from scratch. We propose DECODE, a continual learning framework that starts from a pre-trained generalized model and incrementally expands specialized models for distinct domains. Unlike existing approaches that pursue a single unified model, DECODE explicitly balances specialization and generalization through dynamic model selection. It employs a hypernetwork for parameter generation, which reduces storage costs, and utilizes a normalizing flow for real-time domain inference via likelihood estimation. Outputs from specialized and generalized models are fused using Bayesian uncertainty estimation. This integration ensures optimal performance in familiar conditions while maintaining robustness in novel scenarios. Extensive experiments show DECODE achieves a low forgetting rate of 0.044 and an average minADE of 0.584 m, outperforming prior methods and generalizing well across diverse driving domains. Furthermore, we demonstrate that DECODE can be extended beyond motion prediction to general continual learning tasks such as image classification, showcasing its broad applicability.
6 January 2026 in The AAPS Journal
The phenomenon of nonlinear pharmacokinetics (PK) mediated by a drug’s pharmacological target, also known as target-mediated drug disposition (TMDD), has been increasingly observed in small molecules in the past decade. TMDD class effect with remarkably similar nonlinear PK behaviors has been reported in 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD1) inhibitors, monoamine oxidase type B (MAO-B) inhibitors, and soluble epoxide hydrolase (sEH) inhibitors. We anticipated that the occurrence of TMDD class effect might be due to their target capacities falling within a specific range, where nonlinear PK mediated by target binding are more likely to be evident. To test our hypothesis, we employed a mass spectrometry (MS)-based global proteomics approach to quantify the absolute protein concentrations of 11β-HSD1, sEH, and MAO-B in different tissues across species. The estimated total amounts of 11β-HSD1, MAO-B, and sEH in humans were approximately 4994, 4629, and 4137 nmol, respectively. The comparable abundance levels of these proteins suggest that TMDD is more likely to be observed when a drug binds to a target within a specific range, potentially between 1000 nmol and 10000 nmol, which corresponds to nonlinear PK at doses of 1–10 mg for a compound with a molecular weight of 400 g/mol. Our study highlights the importance of early target quantification and provides valuable insights into predicting unusual nonlinear PK caused by TMDD. Additionally, this proteomics-based approach for quantifying absolute target capacity could serve as a valuable tool for both industry and academic researchers in investigating other pharmacological targets. Graphical Abstract
6 May 2026 in Transportation Research Part A: Policy and Practice