CSpace

浏览/检索结果: 共5条,第1-5条 帮助

限定条件            
已选(0)清除 条数/页:   排序方式:
On the Cybernetics of Crowdsourcing Innovation: A Process Model 期刊论文
IEEE ACCESS, 2022, 卷号: 10, 页码: 27255-27269
作者:  Lin, Lei;  Shen, Huawei;  Liu, Shenghua;  Xu, Li;  Cheng, Xueqi
收藏  |  浏览/下载:18/0  |  提交时间:2022/12/07
Technological innovation  Crowdsourcing  Task analysis  Systematics  Problem-solving  Collective intelligence  Process control  Collaborative work  collective intelligence  crowdsourcing  cybernetics  knowledge management  open innovation  problem-solving  
Feature Rescaling and Fusion for Tiny Object Detection 期刊论文
IEEE ACCESS, 2021, 卷号: 9, 页码: 62946-62955
作者:  Liu, Jingwei;  Gu, Yi;  Han, Shumin;  Zhang, Zhibin;  Guo, Jiafeng;  Cheng, Xueqi
收藏  |  浏览/下载:29/0  |  提交时间:2021/12/01
Feature extraction  Object detection  Semantics  Task analysis  Training  Spatial resolution  Shape  Tiny object detection  nonparametric adaptive selection  feature fusion  feature pyramid network  ensemble model  
Att-FPA: Boosting Feature Perceive for Object Detection 期刊论文
IEEE ACCESS, 2021, 卷号: 9, 页码: 47380-47390
作者:  Liu, Jingwei;  Gu, Yi;  Han, Shumin;  Zhang, Zhibin;  Guo, Jiafeng;  Cheng, Xueqi
收藏  |  浏览/下载:28/0  |  提交时间:2021/12/01
Feature extraction  Location awareness  Task analysis  Object detection  Proposals  Neck  Semantics  Deep learning  computer vision  object detection  attention  feature representation  
Roulette: A Pruning Framework to Train a Sparse Neural Network From Scratch 期刊论文
IEEE ACCESS, 2021, 卷号: 9, 页码: 51134-51145
作者:  Zhong, Qiaoling;  Zhang, Zhibin;  Qiu, Qiang;  Cheng, Xueqi
收藏  |  浏览/下载:34/0  |  提交时间:2021/12/01
Network pruning  inference acceleration  model compression  multiple GPUs  
Trend-Smooth: Accelerate Asynchronous SGD by Smoothing Parameters Using Parameter Trends 期刊论文
IEEE ACCESS, 2019, 卷号: 7, 页码: 156848-156859
作者:  Cui, Guoxin;  Guo, Jiafeng;  Fan, Yixing;  Lan, Yanyan;  Cheng, Xueqi
收藏  |  浏览/下载:35/0  |  提交时间:2020/12/10
Training  Market research  Acceleration  Convergence  Servers  Stochastic processes  Machine learning  Parameter trend  asynchronous SGD  accelerate training