论文常用词汇

读音

hypothesis haɪˈpɒθəsɪs
parameter pəˈræmɪtə(r)
reinforce ˌri:ɪnˈfɔ:s

短语

通用领域

state of the art 最新的进展，简写为SOTA
revive 使复苏，使某技术重新被使用。
orthogonal 正交的
deviation 偏差。例如：Our metric should reflect that small deviations in large boxes matter less than in small boxes
convergence 收敛性。例：batch normalization leads to significant improvements in convergence while eliminating the need for other forms of regularization
fine tune 微调。e.g. we first fine tune the classification network at the full 448*448 resolution for 10 epochs on ImageNet.
on the fly 动态的。e.g. However, since our model only uses convolutional and pooling layers it can be resized on the fly.
on par with 于……相提并论。e.g. It is almost on par with RetinaNet and far above the SSD variants.
auxiliary 辅助的，附加的。e.g. Tiny SSD is composed of a non-uniform highly optimized Fire sub-network stack, which feeds into a non-uniform sub-network stack of highly optimized SSD-based auxiliary convolutional feature layers, designed specifically to minimize model size while retaining object detection performance.
without bells and whistles 没有花哨的东西，没有trick，很直白，健壮的。常出现于Kaiming-He的论文。e.g. Using FPN in a basic Faster R-CNN system, our method achieves SOTA single-model results on the COCO detection benchmark without bells and whistles

图像领域

resolution 分辨率。e.g. The original YOLO trains the classifier network at 224*224 and increases the resolution to 448 for detection.
semantic information 语义信息。e.g. This method allows us to get more meaningful semantic information from the upsampled features and finer grained information from the earlier feature map
microarchitecture 微型结构。比如一个残差块、inception、dense block
macroarchitecture 宏观结构。e.g. Perhaps the mostly widely studied CNN macroarchitecture topic in the recent literature is the impact of depth in networks.
negligible 微不足道的。e.g. For Tiny SSD, the parameter are represented in half precision floating-point, thus leading to further deep neural network model size reductions while having a negligible effect on object detection accuracy.