Sequential Feature Filtering Classifier
This is an incremental improvement for computer vision tasks like classification, detection, segmentation, and action recognition.
The paper tackles the problem of improving convolutional neural network performance across multiple tasks by proposing the Sequential Feature Filtering Classifier (FFC), which uses sequential LayerNorm and ReLU to filter features, resulting in significant performance gains with negligible overhead.
We propose Sequential Feature Filtering Classifier (FFC), a simple but effective classifier for convolutional neural networks (CNNs). With sequential LayerNorm and ReLU, FFC zeroes out low-activation units and preserves high-activation units. The sequential feature filtering process generates multiple features, which are fed into a shared classifier for multiple outputs. FFC can be applied to any CNNs with a classifier, and significantly improves performances with negligible overhead. We extensively validate the efficacy of FFC on various tasks: ImageNet-1K classification, MS COCO detection, Cityscapes segmentation, and HMDB51 action recognition. Moreover, we empirically show that FFC can further improve performances upon other techniques, including attention modules and augmentation techniques. The code and models will be publicly available.