LPRNet: Lightweight Deep Network by Low-rank Pointwise Residual Convolution
This work addresses model compression for deployment on devices like smartphones, representing an incremental improvement in lightweight deep learning methods.
The paper tackles the challenge of deploying deep learning models on resource-constrained devices by proposing LPRNet, a lightweight module using low-rank pointwise residual convolution, which achieves competitive performance on image classification and face alignment benchmarks while significantly reducing FLOPs and memory costs compared to state-of-the-art compressed models.
Deep learning has become popular in recent years primarily due to the powerful computing device such as GPUs. However, deploying these deep models to end-user devices, smart phones, or embedded systems with limited resources is challenging. To reduce the computation and memory costs, we propose a novel lightweight deep learning module by low-rank pointwise residual (LPR) convolution, called LPRNet. Essentially, LPR aims at using low-rank approximation in pointwise convolution to further reduce the module size, while keeping depthwise convolutions as the residual module to rectify the LPR module. This is critical when the low-rankness undermines the convolution process. We embody our design by replacing modules of identical input-output dimension in MobileNet and ShuffleNetv2. Experiments on visual recognition tasks including image classification and face alignment on popular benchmarks show that our LPRNet achieves competitive performance but with significant reduction of Flops and memory cost compared to the state-of-the-art deep models focusing on model compression.