Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userScaling Local Self-Attention for Parameter Efficient Visual Backbones
162
0
0.0
(
0
)
Added by
Ashish Vaswani
Publication date
2021
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
Self-attention has the promise of improving computer vision systems due to parameter-independent scaling of receptive fields and content-dependent interactions, in contrast to parameter-dependent scaling and content-independent interactions of convolutions. Self-attention models have recently been shown to have encouraging improvements on accuracy-parameter trade-offs compared to baseline convolutional models such as ResNet-50. In this work, we aim to develop self-attention models that can outperform not just the canonical baseline models, but even the high-performing convolutional models. We propose two extensions to self-attention that, in conjunction with a more efficient implementation of self-attention, improve the speed, memory usage, and accuracy of these models. We leverage these improvements to develop a new self-attention model family, HaloNets, which reach state-of-the-art accuracies on the parameter-limited setting of the ImageNet classification benchmark. In preliminary transfer learning experiments, we find that HaloNet models outperform much larger models and have better inference performance. On harder tasks such as object detection and instance segmentation, our simple local self-attention and convolutional hybrids show improvements over very strong baselines. These results mark another step in demonstrating the efficacy of self-attention models on settings traditionally dominated by convolutional models.
rate research
Read More
Attention mechanisms, especially self-attention, have played an increasingly important role in deep feature representation for visual tasks. Self-attention updates the feature at each position by computing a weighted sum of features using pair-wise affinities across all positions to capture the long-range dependency within a single sample. However, self-attention has quadratic complexity and ignores potential correlation between different samples. This paper proposes a novel attention mechanism which we call external attention, based on two external, small, learnable, shared memories, which can be implemented easily by simply using two cascaded linear layers and two normalization layers; it conveniently replaces self-attention in existing popular architectures. External attention has linear complexity and implicitly considers the correlations between all data samples. We further incorporate the multi-head mechanism into external attention to provide an all-MLP architecture, external attention MLP (EAMLP), for image classification. Extensive experiments on image classification, object detection, semantic segmentation, instance segmentation, image generation, and point cloud analysis reveal that our method provides results comparable or superior to the self-attention mechanism and some of its variants, with much lower computational and memory costs.
Self-Attention has become prevalent in computer vision models. Inspired by fully connected Conditional Random Fields (CRFs), we decompose it into local and context terms. They correspond to the unary and binary terms in CRF and are implemented by attention mechanisms with projection matrices. We observe that the unary terms only make small contributions to the outputs, and meanwhile standard CNNs that rely solely on the unary terms achieve great performances on a variety of tasks. Therefore, we propose Locally Enhanced Self-Attention (LESA), which enhances the unary term by incorporating it with convolutions, and utilizes a fusion module to dynamically couple the unary and binary operations. In our experiments, we replace the self-attention modules with LESA. The results on ImageNet and COCO show the superiority of LESA over convolution and self-attention baselines for the tasks of image recognition, object detection, and instance segmentation. The code is made publicly available.
We developed a new and computationally simple local block-wise self attention based normal structures segmentation approach applied to head and neck computed tomography (CT) images. Our method uses the insight that normal organs exhibit regularity in their spatial location and inter-relation within images, which can be leveraged to simplify the computations required to aggregate feature information. We accomplish this by using local self attention blocks that pass information between each other to derive the attention map. We show that adding additional attention layers increases the contextual field and captures focused attention from relevant structures. We developed our approach using U-net and compared it against multiple state-of-the-art self attention methods. All models were trained on 48 internal headneck CT scans and tested on 48 CT scans from the external public domain database of computational anatomy dataset. Our method achieved the highest Dice similarity coefficient segmentation accuracy of 0.85$pm$0.04, 0.86$pm$0.04 for left and right parotid glands, 0.79$pm$0.07 and 0.77$pm$0.05 for left and right submandibular glands, 0.93$pm$0.01 for mandible and 0.88$pm$0.02 for the brain stem with the lowest increase of 66.7% computing time per image and 0.15% increase in model parameters compared with standard U-net. The best state-of-the-art method called point-wise spatial attention, achieved textcolor{black}{comparable accuracy but with 516.7% increase in computing time and 8.14% increase in parameters compared with standard U-net.} Finally, we performed ablation tests and studied the impact of attention block size, overlap of the attention blocks, additional attention layers, and attention block placement on segmentation performance.
Visual navigation for autonomous agents is a core task in the fields of computer vision and robotics. Learning-based methods, such as deep reinforcement learning, have the potential to outperform the classical solutions developed for this task; however, they come at a significantly increased computational load. Through this work, we design a novel approach that focuses on performing better or comparable to the existing learning-based solutions but under a clear time/computational budget. To this end, we propose a method to encode vital scene semantics such as traversable paths, unexplored areas, and observed scene objects -- alongside raw visual streams such as RGB, depth, and semantic segmentation masks -- into a semantically informed, top-down egocentric map representation. Further, to enable the effective use of this information, we introduce a novel 2-D map attention mechanism, based on the successful multi-layer Transformer networks. We conduct experiments on 3-D reconstructed indoor PointGoal visual navigation and demonstrate the effectiveness of our approach. We show that by using our novel attention schema and auxiliary rewards to better utilize scene semantics, we outperform multiple baselines trained with only raw inputs or implicit semantic information while operating with an 80% decrease in the agents experience.
Recently, Vision Transformer and its variants have shown great promise on various computer vision tasks. The ability of capturing short- and long-range visual dependencies through self-attention is arguably the main source for the success. But it also brings challenges due to quadratic computational overhead, especially for the high-resolution vision tasks (e.g., object detection). In this paper, we present focal self-attention, a new mechanism that incorporates both fine-grained local and coarse-grained global interactions. Using this new mechanism, each token attends the closest surrounding tokens at fine granularity but the tokens far away at coarse granularity, and thus can capture both short- and long-range visual dependencies efficiently and effectively. With focal self-attention, we propose a new variant of Vision Transformer models, called Focal Transformer, which achieves superior performance over the state-of-the-art vision Transformers on a range of public image classification and object detection benchmarks. In particular, our Focal Transformer models with a moderate size of 51.1M and a larger size of 89.8M achieve 83.5 and 83.8 Top-1 accuracy, respectively, on ImageNet classification at 224x224 resolution. Using Focal Transformers as the backbones, we obtain consistent and substantial improvements over the current state-of-the-art Swin Transformers for 6 different object detection methods trained with standard 1x and 3x schedules. Our largest Focal Transformer yields 58.7/58.9 box mAPs and 50.9/51.3 mask mAPs on COCO mini-val/test-dev, and 55.4 mIoU on ADE20K for semantic segmentation, creating new SoTA on three of the most challenging computer vision tasks.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
