Specifically, robustness with only ACM loss is 48.38%, the addition of soft-labels improves it to 49.53%, the addition of mixup improves it to 52.29%, and the addition of both of these components make final robustness to 56.65%. Also, note that only soft labels are not enough to transfer robustness in this case, as shown by KDOnly column. This is in line with the observations of Goldblum et al. [4]. A.4.2 Role of Intermediate Features To understand the role of low, mid, and high-level features, we performed experiments on CIFAR-10 by progressively changing blocks used for distillation. For this ablation study, we kept all the standard settings reported in the Section A.1. Our correspondence of blocks and features is as follows: block 2: low-level features; block 3: mid-level features; block 4: high-level features. Please note that block 1 corresponds to the output of the first layer only. Therefore, we do not call it low-level features.

Knowledge Distillation (KD) is a well-known training paradigm in deep neural networks where knowledge acquired by a large teacher model is transferred to a small student. KD has proven to be an effective technique to significantly improve the student's performance for various tasks including object detection. As such, KD techniques mostly rely on guidance at the intermediate feature level, which is typically implemented by minimizing an ℓp-norm distance between teacher and student activations during training. In this paper, we propose a replacement for the pixel-wise independent ℓp-norm based on the structural similarity (SSIM) [28]. By taking into account additional contrast and structural cues, feature importance, correlation and spatial dependence in the feature space are considered in the loss formulation. Extensive experiments on MSCOCO [16] demonstrate the effectiveness of our method across different training schemes and architectures. Our method adds only little computational overhead, is straightforward to implement and at the same time it significantly outperforms the standard ℓp-norms. Moreover, more complex state-of-the-art KD methods [13, 33] using attention-based sampling mechanisms are outperformed, including a +3.5 AP gain using a Faster R-CNN R-50 [21] compared to a vanilla model.

Deep learning models have achieved impressive performance on a wide variety of challenging tasks such as image recognition, natural language generation, game playing, etc.

Since pioneering work of Hinton et al., knowledge distillation based on Kullback-Leibler Divergence (KL-Div) has been predominant, and recently its variants have

Recent advancements have shown promise in applying traditional Semi-Supervised Learning strategies to the task of Generalized Category Discovery (GCD).