2022-07-14 更新
Transformer Compressed Sensing via Global Image Tokens
Authors:Marlon Bran Lorenzana, Craig Engstrom, Shekhar S. Chandra
Convolutional neural networks (CNN) have demonstrated outstanding Compressed Sensing (CS) performance compared to traditional, hand-crafted methods. However, they are broadly limited in terms of generalisability, inductive bias and difficulty to model long distance relationships. Transformer neural networks (TNN) overcome such issues by implementing an attention mechanism designed to capture dependencies between inputs. However, high-resolution tasks typically require vision Transformers (ViT) to decompose an image into patch-based tokens, limiting inputs to inherently local contexts. We propose a novel image decomposition that naturally embeds images into low-resolution inputs. These Kaleidoscope tokens (KD) provide a mechanism for global attention, at the same computational cost as a patch-based approach. To showcase this development, we replace CNN components in a well-known CS-MRI neural network with TNN blocks and demonstrate the improvements afforded by KD. We also propose an ensemble of image tokens, which enhance overall image quality and reduces model size. Supplementary material is available: https://github.com/uqmarlonbran/TCS.git
PDF 4 Pages, 4 Figures, 2 Tables
点此查看论文截图
LightViT: Towards Light-Weight Convolution-Free Vision Transformers
Authors:Tao Huang, Lang Huang, Shan You, Fei Wang, Chen Qian, Chang Xu
Vision transformers (ViTs) are usually considered to be less light-weight than convolutional neural networks (CNNs) due to the lack of inductive bias. Recent works thus resort to convolutions as a plug-and-play module and embed them in various ViT counterparts. In this paper, we argue that the convolutional kernels perform information aggregation to connect all tokens; however, they would be actually unnecessary for light-weight ViTs if this explicit aggregation could function in a more homogeneous way. Inspired by this, we present LightViT as a new family of light-weight ViTs to achieve better accuracy-efficiency balance upon the pure transformer blocks without convolution. Concretely, we introduce a global yet efficient aggregation scheme into both self-attention and feed-forward network (FFN) of ViTs, where additional learnable tokens are introduced to capture global dependencies; and bi-dimensional channel and spatial attentions are imposed over token embeddings. Experiments show that our model achieves significant improvements on image classification, object detection, and semantic segmentation tasks. For example, our LightViT-T achieves 78.7% accuracy on ImageNet with only 0.7G FLOPs, outperforming PVTv2-B0 by 8.2% while 11% faster on GPU. Code is available at https://github.com/hunto/LightViT.
PDF 13 pages, 7 figures, 9 tables
点此查看论文截图
TRT-ViT: TensorRT-oriented Vision Transformer
Authors:Xin Xia, Jiashi Li, Jie Wu, Xing Wang, Xuefeng Xiao, Min Zheng, Rui Wang
We revisit the existing excellent Transformers from the perspective of practical application. Most of them are not even as efficient as the basic ResNets series and deviate from the realistic deployment scenario. It may be due to the current criterion to measure computation efficiency, such as FLOPs or parameters is one-sided, sub-optimal, and hardware-insensitive. Thus, this paper directly treats the TensorRT latency on the specific hardware as an efficiency metric, which provides more comprehensive feedback involving computational capacity, memory cost, and bandwidth. Based on a series of controlled experiments, this work derives four practical guidelines for TensorRT-oriented and deployment-friendly network design, e.g., early CNN and late Transformer at stage-level, early Transformer and late CNN at block-level. Accordingly, a family of TensortRT-oriented Transformers is presented, abbreviated as TRT-ViT. Extensive experiments demonstrate that TRT-ViT significantly outperforms existing ConvNets and vision Transformers with respect to the latency/accuracy trade-off across diverse visual tasks, e.g., image classification, object detection and semantic segmentation. For example, at 82.7% ImageNet-1k top-1 accuracy, TRT-ViT is 2.7$\times$ faster than CSWin and 2.0$\times$ faster than Twins. On the MS-COCO object detection task, TRT-ViT achieves comparable performance with Twins, while the inference speed is increased by 2.8$\times$.
PDF
点此查看论文截图
Pyramid Transformer for Traffic Sign Detection
Authors:Omid Nejati Manzari, Amin Boudesh, Shahriar B. Shokouhi
Traffic sign detection is a vital task in the visual system of self-driving cars and the automated driving system. Recently, novel Transformer-based models have achieved encouraging results for various computer vision tasks. We still observed that vanilla ViT could not yield satisfactory results in traffic sign detection because the overall size of the datasets is very small and the class distribution of traffic signs is extremely unbalanced. To overcome this problem, a novel Pyramid Transformer with locality mechanisms is proposed in this paper. Specifically, Pyramid Transformer has several spatial pyramid reduction layers to shrink and embed the input image into tokens with rich multi-scale context by using atrous convolutions. Moreover, it inherits an intrinsic scale invariance inductive bias and is able to learn local feature representation for objects at various scales, thereby enhancing the network robustness against the size discrepancy of traffic signs. The experiments are conducted on the German Traffic Sign Detection Benchmark (GTSDB). The results demonstrate the superiority of the proposed model in the traffic sign detection tasks. More specifically, Pyramid Transformer achieves 75.6% mAP in GTSDB when applied to the Cascade RCNN as the backbone and surpassing most well-known and widely used SOTAs.
PDF
点此查看论文截图