Despite the growing integration of deep models into mobile terminals, the accuracy of these models declines significantly due to various deployment interferences. Test-time adaptation (TTA) has emerged to improve the performance of deep models by adapting them to unlabeled target data online. Yet, the significant memory cost, particularly in resource-constrained terminals, impedes the effective deployment of most backward-propagation-based TTA methods. To tackle memory constraints, we introduce Surgeon, a method that substantially reduces memory cost while preserving comparable accuracy improvements during fully test-time adaptation (FTTA) without relying on specific network architectures or modifications to the original training procedure. Specifically, we propose a novel dynamic activation sparsity strategy that directly prunes activations at layer-specific dynamic ratios during adaptation, allowing for flexible control of learning ability and memory cost in a data-sensitive manner. Among this, two metrics, Gradient Importance and Layer Activation Memory, are considered to determine the layer-wise pruning ratios, reflecting accuracy contribution and memory efficiency, respectively. Experimentally, our method surpasses the baselines by not only reducing memory usage but also achieving superior accuracy, delivering SOTA performance across diverse datasets, architectures, and tasks.

Neural enhancement through super-resolution (SR) deep neural networks (DNNs) opens up new possibilities for ultra-high-definition (UHD) live streaming. Yet, the heavy SR DNN inference overhead leads to severe deployment challenges. To reduce the overhead, existing systems propose to apply DNN-based SR only on carefully selected anchor frames while upscaling non-anchor frames via the lightweight reusing-based SR approach. However, frame-level scheduling is coarse-grained and fails to deliver optimal efficiency. In this work, we propose Palantír, the first neural-enhanced UHD live streaming system with fine-grained patch-level scheduling.

In recent years, large language models (LLMs) have shown rapid development, becoming one of the most popular topics in the field of artificial intelligence. LLMs have demonstrated powerful generalization and learning capabilities, and their performance on various language tasks has been remarkable. Despite their successes, LLMs face significant challenges, particularly in domain-specific tasks that require structured knowledge, often leading to issues such as hallucinations. To mitigate these challenges, we propose a novel system, SynaptiQA, which integrates LLMs with Knowledge Graphs (KGs) to answer more questions about knowledge. Our approach leverages the generative capabilities of LLMs to create and optimize KG queries, thereby improving the accuracy and contextual relevance of responses. Experimental results in an industrial data set demonstrate that SynaptiQA outperforms baseline models and naive retrieval-augmented generation (RAG) systems, demonstrating improved accuracy and reduced hallucinations. This integration of KGs with LLMs paves the way for more reliable and interpretable domain-specific question answering systems.

The digital transformation of factories has greatly increased the number of peripherals that need to connect to a network for sensing or control, resulting in a growing demand for a new network category known as the Equipment Area Network (EAN). The EAN is characterized by its cable-free, high-capacity, low-latency, and low-power features. To meet these expectations, we presentBEANet, a novel solution designed specifically for EAN that combines a two-stage synchronization mechanism with a time division protocol. We implemented the system using commercially available Bluetooth Low Energy (BLE) modules and evaluated its performance. Our results show that the network can support up to 150 peripherals with a packet reception rate of 95.4%, which is only 0.9% lower than collision-free BLE transmission. When the cycle time is set to 2 s, the average transmission latency for all peripherals is 0.1 s, while the power consumption is 18.9 μW, which is only half that of systems using LLDN or TSCH. Simulation results also demonstrate that BEANet has the potential to accommodate over 30,000 peripherals under certain configurations.

LoRa and its enabled LoRa wide-area network (LoRaWAN) have been seen as an important part of the next-generation network for massive Internet-of-Things (IoT). Due to LoRa's low-power and long-range nature, LoRa signals are much weaker than the noise floor, particularly in complex urban or semi-indoor environments. Therefore, weak signal decoding is critical to achieve the desired wide-area coverage in general. Existing work has shown the advantages of exploring deep neural networks (DNN) for weak signal decoding. However, the existing single-gateway based DNN decoder is hard to fully leverage the spatial information in multi-gateway scenarios. In this paper, we propose SRLoRa, an efficient DNN LoRa decoder that fully utilizes the spatial information from multiple gateways to decode extremely weak LoRa signals. Specifically, we design interleaving denoising and merging layers to improve signal quality at ultra-low SNR. We develop efficient merging on feature maps extracted by denoising DNNs to tolerate time misalignments among different signals. We define max and min operations in the merging layer to efficiently extract salient features and reduce noise, merging the features extracted from multiple gateways to guide future DNN layers to gradually improve signal quality. We implement SRLoRa with USPR N210 and commercial LoRa nodes and evaluate its performance indoors and outdoors. The results show that with four gateways, SRLoRa achieves SNR gain at 4.53--4.82 dB, which is 2.51× of Charm, leading to a 1.84× coverage area compared to standard LoRa in an urban deployment.