# SBNet: Sparse Blocks Network for Fast Inference

arXiv: https://arxiv.org/abs/1801.02108 · HTML: https://mlsanity.com/paper/1801.02108/

## Key facts

- Cited by 54 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 Jan 2018.

## Built on by (most cited first, 30 of 54)

- [Dynamic Neural Networks: A Survey](https://mlsanity.com/paper/2102.04906/) (2021)
- [Focal Sparse Convolutional Networks for 3D Object Detection](https://mlsanity.com/paper/2204.12463/) (2022)
- [FlowFormer++: Masked Cost Volume Autoencoding for Pretraining Optical Flow Estimation](https://mlsanity.com/paper/2303.01237/) (2023)
- [ConvMAE: Masked Convolution Meets Masked Autoencoders](https://mlsanity.com/paper/2205.03892/) (2022)
- [DistriFusion: Distributed Parallel Inference for High-Resolution Diffusion Models](https://mlsanity.com/paper/2402.19481/) (2024)
- [QueryDet: Cascaded Sparse Query for Accelerating High-Resolution Small Object Detection](https://mlsanity.com/paper/2103.09136/) (2021)
- [Sparse Tensor-based Multiscale Representation for Point Cloud Geometry Compression](https://mlsanity.com/paper/2111.10633/) (2021)
- [SparseViT: Revisiting Activation Sparsity for Efficient High-Resolution Vision Transformer](https://mlsanity.com/paper/2303.17605/) (2023)
- [Radar-Camera Fusion for Object Detection and Semantic Segmentation in Autonomous Driving: A Comprehensive Review](https://mlsanity.com/paper/2304.10410/) (2023)
- [A Dynamic Multi-Scale Voxel Flow Network for Video Prediction](https://mlsanity.com/paper/2303.09875/) (2023)
- [PnP-DETR: Towards Efficient Visual Analysis with Transformers](https://mlsanity.com/paper/2109.07036/) (2021)
- [ZeroFL: Efficient On-Device Training for Federated Learning with Local Sparsity](https://mlsanity.com/paper/2208.02507/) (2022)
- [Efficient Spatially Sparse Inference for Conditional GANs and Diffusion Models](https://mlsanity.com/paper/2211.02048/) (2022)
- [AdaFocus V2: End-to-End Training of Spatial Dynamic Networks for Video Recognition](https://mlsanity.com/paper/2112.14238/) (2021)
- [Skip-Convolutions for Efficient Video Processing](https://mlsanity.com/paper/2104.11487/) (2021)
- [Object-Centric Diffusion for Efficient Video Editing](https://mlsanity.com/paper/2401.05735/) (2024)
- [Dual-side Sparse Tensor Core](https://mlsanity.com/paper/2105.09564/) (2021)
- [Spatial Pruned Sparse Convolution for Efficient 3D Object Detection](https://mlsanity.com/paper/2209.14201/) (2022)
- [Latency-aware Unified Dynamic Networks for Efficient Image Recognition](https://mlsanity.com/paper/2308.15949/) (2023)
- [Latency-aware Spatial-wise Dynamic Networks](https://mlsanity.com/paper/2210.06223/) (2022)
- [Prompt-Driven Dynamic Object-Centric Learning for Single Domain Generalization](https://mlsanity.com/paper/2402.18447/) (2024)
- [Exploring Radar Data Representations in Autonomous Driving: A Comprehensive Review](https://mlsanity.com/paper/2312.04861/) (2023)
- [PFA: Privacy-preserving Federated Adaptation for Effective Model Personalization](https://mlsanity.com/paper/2103.01548/) (2021)
- [Anytime Dense Prediction with Confidence Adaptivity](https://mlsanity.com/paper/2104.00749/) (2021)
- [DeltaCNN: End-to-End CNN Inference of Sparse Frame Differences in Videos](https://mlsanity.com/paper/2203.03996/) (2022)
- [Sparse PointPillars: Maintaining and Exploiting Input Sparsity to Improve Runtime on Embedded Systems](https://mlsanity.com/paper/2106.06882/) (2021)
- [Accelerating Text-to-Image Editing via Cache-Enabled Sparse Diffusion Inference](https://mlsanity.com/paper/2305.17423/) (2023)
- [Towards Efficient Use of Multi-Scale Features in Transformer-Based Object Detectors](https://mlsanity.com/paper/2208.11356/) (2022)
- [Accelerating Deep Neural Networks via Semi-Structured Activation Sparsity](https://mlsanity.com/paper/2309.06626/) (2023)
- [CrossRoI: Cross-camera Region of Interest Optimization for Efficient Real Time Video Analytics at Scale](https://mlsanity.com/paper/2105.06524/) (2021)

## Abstract (arXiv)

Conventional deep convolutional neural networks (CNNs) apply convolution operators uniformly in space across all feature maps for hundreds of layers - this incurs a high computational cost for real-time applications. For many problems such as object detection and semantic segmentation, we are able to obtain a low-cost computation mask, either from a priori problem knowledge, or from a low-resolution segmentation network. We show that such computation masks can be used to reduce computation in the high-resolution main network. Variants of sparse activation CNNs have previously been explored on small-scale tasks and showed no degradation in terms of object classification accuracy, but often measured gains in terms of theoretical FLOPs without realizing a practical speed-up when compared to highly optimized dense convolution implementations. In this work, we leverage the sparsity structure of computation masks and propose a novel tiling-based sparse convolution algorithm. We verified the effectiveness of our sparse CNN on LiDAR-based 3D object detection, and we report significant wall-clock speed-ups compared to dense convolution without noticeable loss of accuracy.