Search for: All records

The COVID-19 pandemic has intensified the need for home-based cardiac health monitoring systems. Despite advancements in electrocardiograph (ECG) and phonocardiogram (PCG) wearable sensors, accurate heart sound segmentation algorithms remain understudied. Existing deep learning models, such as convolutional neural networks (CNN) and recurrent neural networks (RNN), struggle to segment noisy signals using only PCG data. We propose a two-step heart sound segmentation algorithm that analyzes synchronized ECG and PCG signals. The first step involves heartbeat detection using a CNN-LSTM-based model on ECG data, and the second step focuses on beat-wise heart sound segmentation with a 1D U-Net that incorporates multi-modal inputs. Our method leverages temporal correlation between ECG and PCG signals to enhance segmentation performance. To tackle the label-hungry issue in AI-supported biomedical studies, we introduce a segment-wise contrastive learning technique for signal segmentation, overcoming the limitations of traditional contrastive learning methods designed for classification tasks. We evaluated our two-step algorithm using the PhysioNet 2016 dataset and a private dataset from Bayland Scientific, obtaining a 96.43 F1 score on the former. Notably, our segment-wise contrastive learning technique demonstrated effective performance with limited labeled data. When trained on just 1% of labeled PhysioNet data, the model pre-trained on the full unlabeled dataset only dropped 2.88 in the F1 score, outperforming the SimCLR method. Overall, our proposed algorithm and learning technique present promise for improving heart sound segmentation and reducing the need for labeled data. 

The unprecedented success of artificial intelligence (AI) enriches machine learning (ML)-based applications. The availability of big data and compute-intensive algorithms empowers versatility and high accuracy in ML approaches. However, the data processing and innumerable computations burden conventional hardware systems with high power consumption and low performance. Breaking away from the traditional hardware design, non-conventional accelerators exploiting emerging technology have gained significant attention with a leap forward since the emerging devices enable processing-in-memory (PIM) designs of dramatic improvement in efficiency. This paper presents a summary of state-of-the-art PIM accelerators over a decade. The PIM accelerators have been implemented for diverse models and advanced algorithm techniques across diverse neural networks in language processing and image recognition to expedite inference and training. We will provide the implemented designs, methodologies, and results, following the development in the past years. The promising direction of the PIM accelerators, vertically stacking for More than Moore, is also discussed. 

Natural Adversarial Examples (NAEs), images arising naturally from the environment and capable of deceiving classifiers, are instrumental in robustly evaluating and identifying vulnerabilities in trained models. In this work, unlike prior works that passively collect NAEs from real images, we propose to actively synthesize NAEs using the state-of-the-art Stable Diffusion. Specifically, our method formulates a controlled optimization process, where we perturb the token embedding that corresponds to a specified class to generate NAEs. This generation process is guided by the gradient of loss from the target classifier, ensuring that the created image closely mimics the ground-truth class yet fools the classifier. Named SD-NAE (Stable Diffusion for Natural Adversarial Examples), our innovative method is effective in producing valid and useful NAEs, which is demonstrated through a meticulously designed experiment. Code is available at https://github.com/linyueqian/SD-NAE. 

Approximate nearest neighbor search (ANNS) is a key retrieval technique for vector database and many data center applications, such as person re-identification and recommendation systems. It is also fundamental to retrieval augmented generation (RAG) for large language models (LLM) now. Among all the ANNS algorithms, graph-traversal-based ANNS achieves the highest recall rate. However, as the size of dataset increases, the graph may require hundreds of gigabytes of memory, exceeding the main memory capacity of a single workstation node. Although we can do partitioning and use solid-state drive (SSD) as the backing storage, the limited SSD I/O bandwidth severely degrades the performance of the system. To address this challenge, we present NDSEARCh, a hardware-software co-designed near-data processing (NDP) solution for ANNS processing. NDSeARCH consists of a novel in-storage computing architecture, namely, SEARSSD, that supports the ANNS kernels and leverages logic unit (LUN)-level parallelism inside the NAND flash chips. NDSEARCH also includes a processing model that is customized for NDP and cooperates with SearSSD. The processing model enables us to apply a two-level scheduling to improve the data locality and exploit the internal bandwidth in NDSearch, and a speculative searching mechanism to further accelerate the ANNS workload. Our results show that NDSEARCH improves the throughput by up to 31.7×,14.6×,7.4×, and 2.9× over CPU, GPU, a state-of-the-art SmartSSD-only design, and DeepStore, respectively. NDSEARCH also achieves two orders-of-magnitude higher energy efficiency than CPU and GPU.