Effective management of high-resolution, and spatially wide contextual cues is fundamental to the accurate semantic segmentation. Traditional approaches like multi-resolution feature maps, and skip-connection are effe...
详细信息
ISBN:
(数字)9798331531539
ISBN:
(纸本)9798331531546
Effective management of high-resolution, and spatially wide contextual cues is fundamental to the accurate semantic segmentation. Traditional approaches like multi-resolution feature maps, and skip-connection are effective but require changes in the backbone architecture, restricting utilization of newer models and architectures for the problem. In this work we propose an architecture-agnostic, two-stage, global-local frame-work, called GoLo, for the semantic segmentation, which can use arbitrary semantic segmentation models within its two stages. We focus on segmenting cell nuclei in histopathology image analysis, where accurate segmentation of cell nuclei boundaries is one of the key issues. The proposed framework consists of first stage with Global and second stage with Local learning approach. The first stage is proposed to process the image globally and provide the coarse nuclei segmentation map. In the second stage, to process the image locally, coarse segmentation map and input image is first converted into patches. These patches are then fed as input to the second stage to get the fine- grained segmentation map. Both stages are trained with a combination of dice and binary cross entropy loss. To show the effectiveness of our approach, we test 4 state-of-the-art segmentation architectures (ACC-UNet, UCTransnet, Swin-UNet, and Vanilla U-Net), on 4 different benchmark datasets (MoNuSeg, CPM-17, CoNSep, and TNBC). We evaluate performance of each technique before and after using our framework. We report an average improvement of 4.82 % in mIoU, and 4.52% mDSC score, across techniques, and datasets.
Online Social Networking sites have become a well-known way for web surfers to connect and meet. Twitter got to be a well-known micro blogging site that clients post and associate with messages known as tweets. As thi...
详细信息
This article explores the emerging field of in-memory computing, which has the potential to significantly improve energy efficiency in many signal processing and machine learning applications. In-memory computing syst...
详细信息
作者：
Gu, YuxianKe, PeiZhu, XiaoyanHuang, MinlieThe CoAI group
Tsinghua University Beijing China Institute for Artificial Intelligence State Key Lab of Intelligent Technology and Systems Beijing National Research Center for Information Science and Technology Department of Computer Science and Technology Tsinghua University Beijing China
Training language models to learn from human instructions for zero-shot cross-task generalization has attracted much attention in NLP communities. Recently, instruction tuning (IT), which fine-tunes a pre-trained lang...
详细信息
Human brain possesses an extraordinary ability to attend to a specific sound source in a multi-talk, noisy environment such as a cocktail party. Auditory attention detection (AAD) aims to automatically identify such a...
详细信息
In today’s scenario, computer vision is one of the fundamental research areas of artificial intelligence including object detection and object tracking which are the upcoming trends. In the present work, the TransTra...
详细信息
In this paper, a decentralized relay-based approach (D-MRFTE) for unknown area exploration using a team of autonomous mobile robots is proposed under communication constraints. Using the relay robots, the multi-robot ...
In this paper, a decentralized relay-based approach (D-MRFTE) for unknown area exploration using a team of autonomous mobile robots is proposed under communication constraints. Using the relay robots, the multi-robot system forms a high-latency decentralized network with distributed copies of exploration information for which eventual consistency and completeness are ensured through meetups. The meetups act as a safety net and set a bound on latency by ensuring data transfer at periodic intervals whenever the multi-robot network gets fragmented. The information exchange related to the robot’s state and the ongoing exploration is facilitated by the relay robots. The robots use timestamps to assimilate the latest available information by using version vectors. To achieve a consistent state of explorer robots, the relays schedule meetups with other relays they come in contact with, creating a tightly-knit group. Our approach, under two communication models, i.e., Disk-based and Line-of-Sight-based, exhibits superior performance compared with two state-of-the-art algorithms in terms of completion time and distance traveled by the robot team. The simulations are conducted in a Player/Stage simulator with different robot team sizes.
Asia’s Human-computer Interaction (HCI) landscape is rapidly evolving, yet it faces distinct challenges in curriculum development, research establishment, and career navigation. This panel discussion, hosted by Asia ...
详细信息
With the progression of technology, the automobile industry has experienced a significant transformation due to the emergence of Electric Vehicles (EVs). EVs offer a sustainable means of transportation that doesn'...
详细信息
暂无评论