PAPERS

过滤器相应归一化层：在深度神经网络的训练中消除批次依赖性

Deep Learning

A Multigrid Method for Efficiently Training Video Models

Computer Vision and Pattern Recognition

KernelNet:用于深度生成建模的数据依赖的内核参数化

Machine Learning

AP-Perf: 在可微学习中整合通用的性能指标

Deep Learning

高效的卷积神经网络用于基于深度的多姿态估计

Computer Vision and Pattern Recognition

Conclusion-Supplement Answer Generation for Non-Factoid Questions

Natural Language Processing

多域图像分割的对抗性归一化

Machine Learning

通过可赋予的反例实现深度神经网络指纹识别

Deep Learning