Edge deployment of large Vision-Language Models (VLMs) increasingly relies on flash-based weight offloading, where activation sparsification is used to reduce I/O overhead. However, conventional sparsification remains model-centric, selecting neurons solely by activation magnitude and neglecting how access patterns influence flash performance. We present Neuron Chunking, an I/O-efficient sparsification strategy that operates on *chunks*—groups of contiguous neurons in memory—and couples neuron importance with storage access cost. The method models I/O latency through a lightweight abstraction of access contiguity and selects chunks with high utility, defined as neuron importance normalized by estimated latency. By aligning sparsification decisions with the underlying storage behavior, Neuron Chunking improves I/O efficiency by up to 4.65× and 5.76× on Jetson Orin Nano and Jetson AGX Orin, respectively. The code is available at https://github.com/snuhcs/vlm-flash.

Accurate tracking of hand and object poses is essential for immersive XR applications. However, existing methods often struggle with the stringent requirements of XR environments. We present XRMan, a real-time hand-object pose tracking system designed for in-the-wild scenarios. XRMan uses a drift monitoring module for consistent accuracy, while farthest-point sampling and collider-based optimization streamline iterative optimization and reduce latency. Our system reduces end-to-end latency by 38.9% compared to the baseline, with further improvements expected from the collider-based post-optimization module.

Mixed reality devices with near-eye displays unlock new possibilities for innovation and user experiences. Mixed reality applications require a new unified framework that enables seamless analysis of the real world and simulation of realistic virtual content. Designing such a framework faces various challenges, including huge programming efforts of analysis and simulation pipelines, and inconsistencies between real-world and virtual content caused by end-to-end processes across pipelines.

The recent development of DNN and hardware has created new opportunities for mixed-reality applications. These applications demand the ability to analyze the real world and simulate realistic virtual content. However, designing mixed-reality applications faces diverse challenges due to the absence of a unified framework, such as huge programming effort and inconsistencies between the real scene and virtual content induced by end-to-end latency.

Deep learning models have become increasingly prevalent in various domains, necessitating their deployment on resource-constrained devices. Quantization is a promising way to reduce the model complexity in that it keeps model architecture intact and enables the model to operate on specialized hardwares(e.g., NPU, DSP). Input resolution is also essential in making a trade-off between accuracy and computation.