PerSense: Personalized Instance Segmentation in Dense Images
Authors:
Muhammad Ibraheem Siddiqui,
Muhammad Umer Sheikh,
Hassan Abid,
Muhammad Haris Khan
Abstract:
Leveraging large-scale pre-training, vision foundational models showcase notable performance benefits. While recent years have witnessed significant advancements in segmentation algorithms, existing models still face challenges to automatically segment personalized instances in dense and crowded scenarios. The primary factor behind this limitation stems from bounding box-based detections, which ar…
▽ More
Leveraging large-scale pre-training, vision foundational models showcase notable performance benefits. While recent years have witnessed significant advancements in segmentation algorithms, existing models still face challenges to automatically segment personalized instances in dense and crowded scenarios. The primary factor behind this limitation stems from bounding box-based detections, which are constrained by occlusions, background clutter, and object orientation, particularly when dealing with dense images. To this end, we propose PerSense, an end-to-end, training-free, and model-agnostic one-shot framework to address the personalized instance segmentation in dense images. Towards developing this framework, we make following core contributions. (a) We propose an Instance Detection Module (IDM) and leverage a Vision-Language Model, a grounding object detector, and a few-shot object counter (FSOC) to realize a new baseline. (b) To tackle false positives within candidate point prompts, we design Point Prompt Selection Module (PPSM). Both IDM and PPSM transform density maps from FSOC into personalized instance-level point prompts for segmentation and offer a seamless integration in our model-agnostic framework. (c) We introduce a feedback mechanism which enables PerSense to harness the full potential of FSOC by automating the exemplar selection process. (d) To promote algorithmic advances and effective tools for this relatively underexplored task, we introduce PerSense-D, a dataset exclusive to personalized instance segmentation in dense images. We validate the effectiveness of PerSense on the task of personalized instance segmentation in dense images on PerSense-D and comparison with SOTA. Additionally, our qualitative findings demonstrate the adaptability of our framework to images captured in-the-wild.
△ Less
Submitted 22 May, 2024;
originally announced May 2024.
A Secured Health Care Application Architecture for Cyber-Physical Systems
Authors:
Jin Wang,
Hassan Abid,
Sungyoung Lee,
Lei Shu,
Feng Xia
Abstract:
Cyber-physical systems (CPS) can be viewed as a new generation of systems with integrated control, communication and computational capabilities. Like the internet transformed how humans interact with one another, cyber-physical systems will transform how people interact with the physical world. Currently, the study of CPS is still in its infancy and there exist many research issues and challenges…
▽ More
Cyber-physical systems (CPS) can be viewed as a new generation of systems with integrated control, communication and computational capabilities. Like the internet transformed how humans interact with one another, cyber-physical systems will transform how people interact with the physical world. Currently, the study of CPS is still in its infancy and there exist many research issues and challenges ranging from electricity power, health care, transportation and smart building etc. In this paper, an introduction of CPeSC3 (cyber physical enhanced secured wireless sensor networks (WSNs) integrated cloud computing for u-life care) architecture and its application to the health care monitoring and decision support systems is given. The proposed CPeSC3 architecture is composed of three main components, namely 1) communication core, 2) computation core, and 3) resource scheduling and management core. Detailed analysis and explanation are given for relevant models such as cloud computing, real time scheduling and security models. Finally, a medical health care application scenario is presented based on our practical test-bed which has been built for 3 years.
△ Less
Submitted 30 December, 2011;
originally announced January 2012.
