The strong correlation between neurons or filters can significantly weaken the generalization ability of neural networks. Inspired by the well-known Tammes problem, we propose a novel diversity regularization method to address this issue, which makes the normalized weight vectors of neurons or filters distributed on a hypersphere as uniformly as possible, through maximizing the minimal pairwise angles (MMA). This method can easily exert its effect by plugging the MMA regularization term into the loss function with negligible computational overhead. The MMA regularization is simple, efficient, and effective. Therefore, it can be used as a basic regularization method in neural network training. Extensive experiments demonstrate that MMA regularization is able to enhance the generalization ability of various modern models and achieves considerable performance improvements on CIFAR100 and TinyImageNet datasets. In addition, experiments on face verification show that MMA regularization is also effective for feature learning.

Weaknesses: The regularization coefficient is important. It can be seen from Figure 3 that the influences of different coefficients are unstable in a single experiment. To better clarify the influences, the author should report the average results of multiple experiments under the same setting. I wonder if the convergence speed and stability are changed after introducing the MMA regularization in the training process. Therefore, it would be better to give the training loss curves with and without MMA regularization.

The paper has initially received mixed reviews, but post-rebuttal the expert reviewers have converged to the decision that the paper is above the acceptance threshold and that the proposed regularization is of wide interest to the community. The authors are encouraged to incorporate the extra experimental results from the rebuttal into the final version of the paper. Also, the related work section should be revised by incorporating relevant works pointed by the reviewers (and as promised in the rebuttal).

The strong correlation between neurons or filters can significantly weaken the generalization ability of neural networks. Inspired by the well-known Tammes problem, we propose a novel diversity regularization method to address this issue, which makes the normalized weight vectors of neurons or filters distributed on a hypersphere as uniformly as possible, through maximizing the minimal pairwise angles (MMA). This method can easily exert its effect by plugging the MMA regularization term into the loss function with negligible computational overhead. The MMA regularization is simple, efficient, and effective. Therefore, it can be used as a basic regularization method in neural network training.