We propose a smooth regularization technique that instills a strong temporal inductive bias in video recognition models, particularly benefiting lightweight architectures. Our method encourages smoothness in the intermediate-layer embeddings of consecutive frames by modeling their changes as a Gaussian Random Walk (GRW). This penalizes abrupt representational shifts, thereby promoting low-acceleration solutions that better align with the natural temporal coherence inherent in videos. By leveraging this enforced smoothness, lightweight models can more effectively capture complex temporal dynamics. Applied to such models, our technique yields a 3.8% to 6.4% accuracy improvement on Kinetics-600. Notably, the MoViNets model family trained with our smooth regularization improves the current state of the art by 3.8% to 6.1% within their respective FLOP constraints, while MobileNetV3 and the MoViNets-Stream family achieve gains of 4.9% to 6.4% over prior state-of-the-art models with comparable memory footprints. Our code and models are available at https://github.com/cmusatyalab/grw-smoothing.

翻译：我们提出一种平滑正则化技术，该技术为视频识别模型（尤其是轻量级架构）注入了强大的时间归纳偏置。我们的方法通过将连续帧中间层嵌入的变化建模为高斯随机游走，从而鼓励其平滑性。这惩罚了表征的突变，进而促进了低加速度解，这些解能更好地与视频固有的自然时间连贯性保持一致。通过利用这种强制的平滑性，轻量级模型能够更有效地捕捉复杂的时间动态。将本技术应用于此类模型后，在Kinetics-600数据集上获得了3.8%至6.4%的准确率提升。值得注意的是，采用我们平滑正则化训练的MoViNets模型系列，在各自FLOP约束下将当前最佳性能提升了3.8%至6.1%；而MobileNetV3和MoViNets-Stream模型系列在具有可比内存占用的前提下，相较于先前的最佳模型取得了4.9%至6.4%的性能增益。我们的代码和模型已发布于https://github.com/cmusatyalab/grw-smoothing。