Tied Block Convolution: Leaner and Better CNNs with Shared Thinner Filters
This work addresses inefficiencies in CNNs for computer vision tasks, offering a method to reduce redundancy and improve performance in occluded scenarios, though it is incremental as it builds on existing convolution techniques.
The paper tackles the problem of filter correlation in CNNs by proposing Tied Block Convolution (TBC), which shares thinner filters across channel blocks, resulting in a 6% average precision gain for object detection on MS-COCO under 80% occlusion and up to 64 times parameter reduction in attention modules.
Convolution is the main building block of convolutional neural networks (CNN). We observe that an optimized CNN often has highly correlated filters as the number of channels increases with depth, reducing the expressive power of feature representations. We propose Tied Block Convolution (TBC) that shares the same thinner filters over equal blocks of channels and produces multiple responses with a single filter. The concept of TBC can also be extended to group convolution and fully connected layers, and can be applied to various backbone networks and attention modules. Our extensive experimentation on classification, detection, instance segmentation, and attention demonstrates TBC's significant across-the-board gain over standard convolution and group convolution. The proposed TiedSE attention module can even use 64 times fewer parameters than the SE module to achieve comparable performance. In particular, standard CNNs often fail to accurately aggregate information in the presence of occlusion and result in multiple redundant partial object proposals. By sharing filters across channels, TBC reduces correlation and can effectively handle highly overlapping instances. TBC increases the average precision for object detection on MS-COCO by 6% when the occlusion ratio is 80%. Our code will be released.