Sparse deep learning has significantly reduced computational costs; however, its irregular distribution of non-zero data complicates data flow, limits data reuse, and increases on-chip SRAM access, thereby elevating chip power consumption. To address these challenges, this work maximizes data reuse to minimize SRAM access through two approaches. First, we propose Effective Index Matching (EIM), which efficiently searches and arranges non-zero operations from compressed data. Second, we propose Shared Index Data Reuse (SIDR), which coordinates operations across Processing Elements (PEs) to regularize their SRAM data access, thereby enabling all data to be reused efficiently. Our approach reduces the access of the SRAM buffer by 86% when compared to the previous design, SparTen. As a result, our design achieves a 2.5× improvement in power efficiency compared to state-of-the-art methods while maintaining a simpler dataflow.

Safety is a critical component of autonomous systems and remains a challenge for learning-based policies to be utilized in the real world. In particular, policies learned using reinforcement learning often fail to generalize to novel environments due to unsafe behavior. In this paper, we propose Sim-to-Lab-to-Real to bridge the reality gap with a probabilistically guaranteed safety-aware policy distribution. To improve safety, we apply a dual policy setup where a performance policy is trained using the cumulative task reward and a backup (safety) policy is trained by solving the Safety Bellman Equation based on Hamilton-Jacobi (HJ) reachability analysis. In Sim-to-Lab transfer, we apply a supervisory control scheme to shield unsafe actions during exploration; in Lab-to-Real transfer, we leverage the Probably Approximately Correct (PAC)-Bayes framework to provide lower bounds on the expected performance and safety of policies in unseen environments. Additionally, inheriting from the HJ reachability analysis, the bound accounts for the expectation over the worst-case safety in each environment. We empirically study the proposed framework for ego-vision navigation in two types of indoor environments with varying degrees of photorealism. We also demonstrate strong generalization performance through hardware experiments in real indoor spaces with a quadrupedal robot. See https://sites.google.com/princeton.edu/sim-to-lab-to-real for supplementary material.

This work examines a novel wireless knowledge caching framework where machine learning models (i.e., knowledge) are cached at local small cell base-stations (SBSs) to facilitate both federated training and access of the models by users. We first consider a single-SBS scenario, where the caching decision, user selection, and wireless resource allocation are jointly determined by minimizing a training error bound subject to constraints on the cache capacity, the communication and computation latency, and the energy consumption. The solution is obtained by first computing the minimum achievable training loss for each model, followed by the optimization of the binary caching variables, which reduces to a 0-1 knapsack problem. The proposed framework is then extended to the multiple-SBS scenario where the user association among SBSs is further examined. We adopt a dual-ascent method where Lagrange multipliers are introduced and updated in each iteration to regularize the dependence among user selection and association. Given the Lagrange multipliers, the caching decision, user selection, resource allocation and user association variables are optimized in turn using a block coordinate descent algorithm. Simulation results show that the proposed scheme can achieve a training error bound that is lower than preference-only and random caching policies in both scenarios.

The ability to anticipate surrounding agents' behaviors is critical to enable safe and seamless autonomous vehicles (AVs). While phenomenological methods have successfully predicted future trajectories from scene context, these predictions lack interpretability. On the other hand, ontological approaches assume an underlying structure able to describe the interaction dynamics or agents' internal decision processes. Still, they often suffer from poor scalability or cannot reflect diverse human behaviors. This work proposes an interpretability framework for a phenomenological method through responsibility evaluations. We formulate responsibility as a measure of how much an agent takes into account the welfare of other agents through counterfactual reasoning. Additionally, this framework abstracts the computed responsibility sequences into different responsibility levels and grounds these latent levels into reward functions. The proposed responsibility-based interpretability framework is modular and easily integrated into a wide range of prediction models. To demonstrate the utility of the proposed framework in providing added interpretability, we adapt an existing AV prediction model and perform a simulation study on a real-world nuScenes traffic dataset. Experimental results show that we can perform offline ex-post traffic analysis by incorporating the responsibility signal and rendering interpretable but accurate online trajectory predictions.

Detailed routing is the most tedious and complex procedure in design automation and has become a determining factor in layout automation in advanced manufacturing nodes. Despite continuing advances in custom integrated circuit (IC) routing research, industrial custom layout flows remain heavily manual due to the high complexity of the custom IC design problem. Besides conventional design objectives such as wirelength minimization, custom detailed routing must also accommodate additional constraints (e.g., path-matching) across the analog/mixed-signal (AMS) and digital domains, making an already challenging procedure even more so. This paper presents a novel detailed routing framework for custom circuits that leverages deep reinforcement learning to optimize routing patterns while considering custom routing constraints and industrial design rules. Comprehensive post-layout analyses based on industrial designs demonstrate the effectiveness of our framework in dealing with the specified constraints and producing sign-off-quality routing solutions.

Deep learning based superresolution achieves high-quality results, but its heavy computational workload, large buffer, and high external memory bandwidth inhibit its usage in mobile devices. To solve the above issues, this paper proposes a real-time hardware accelerator with the tilted layer fusion method that reduces the external DRAM bandwidth by 92% and just needs 102KB on-chip memory. The design implemented with a 40nm CMOS process achieves 1920xl080@60fps throughput with 544. 3K gate count when running at 600MHz; it has higher throughput and lower area cost than previous designs.

In recent years, the rapid development of trade and traffic in our country has resulted in the increasing occurrence of various types of disasters and emergency injuries. According to the statistics from Fire Departments of Taiwan, medical emergency dispatch frequencies have exceeded 300,000, and the frequency of medical emergency dispatches has increased each year. These figures reflect the phenomenon that, despite the continual growth and progression of the community, various accidents and unforeseen situations are becoming increasingly common. Therefore, fire units have established a medical emergency care system with the objective of minimizing injuries, patient suffering, and the risk of mortality, reducing social medical costs, and alleviating the financial burden of medical emergencies on families. Many countries have special medical emergency dispatch systems, but these do not meet Taiwan's rules and emergency procedures. Thus, our research has focused on compiling a symptom-based medical emergency dispatch guide designed to match Taiwan's medical emergency dispatch system. We collate national medical emergency dispatch guides and present a medical emergency dispatch system to support medical staff and to educate these staff on the correct methodology in a short time. This system achieves the following objectives: (1) to enhance the quality of medical emergency services in Taiwan, (2) to reduce the response time of medical emergency dispatch, (3) to reduce the unnecessary waste of medical resources.