一直在等你 一直会等你 计算机视觉中的文章

1、Group Normalization： https://arxiv.org/abs/1803.08494
        解决Batch Normalization中对Batch Size依赖的短板，在目标检测，图像分割，视频分类等任务上，Batch Size往往比较小，导致BN作用的效果比较差。如下图，Group Normalization是对Layer Normalization和Instance Normalization的折中。

def GroupNorm(x, gamma, beta, G, eps=1e-5):
    # x: input features with shape [N,C,H,W]
    # gamma, beta: scale and offset, with shape [1,C,1,1]
    # G: number of groups for GN
    
    N, C, H, W = x.shape
    x = tf.reshape(x, [N, G, C // G, H, W])
    
    mean, var = tf.nn.moments(x, [2, 3, 4], keep dims=True)
    x = (x - mean) / tf.sqrt(var + eps)
    
    x = tf.reshape(x, [N, C, H, W])
    return x * gamma + beta

        根据实验结果，GN比BN对于Batch Size的敏感性更弱，即鲁棒性更高

https://github.com/shaohua0116/Group-Normalization-Tensorflow
2、Focal Loss for Dense Object Detection

3、