ترغب بنشر مسار تعليمي؟ اضغط هنا

GLiT: Neural Architecture Search for Global and Local Image Transformer

295   0   0.0 ( 0 )
 نشر من قبل Peixia Li
 تاريخ النشر 2021
  مجال البحث الهندسة المعلوماتية
والبحث باللغة English




اسأل ChatGPT حول البحث

We introduce the first Neural Architecture Search (NAS) method to find a better transformer architecture for image recognition. Recently, transformers without CNN-based backbones are found to achieve impressive performance for image recognition. However, the transformer is designed for NLP tasks and thus could be sub-optimal when directly used for image recognition. In order to improve the visual representation ability for transformers, we propose a new search space and searching algorithm. Specifically, we introduce a locality module that models the local correlations in images explicitly with fewer computational cost. With the locality module, our search space is defined to let the search algorithm freely trade off between global and local information as well as optimizing the low-level design choice in each module. To tackle the problem caused by huge search space, a hierarchical neural architecture search method is proposed to search the optimal vision transformer from two levels separately with the evolutionary algorithm. Extensive experiments on the ImageNet dataset demonstrate that our method can find more discriminative and efficient transformer variants than the ResNet family (e.g., ResNet101) and the baseline ViT for image classification.



قيم البحث

اقرأ أيضاً

95 - Dong Zhao , Jia Li , Hongyu Li 2021
Recently, the Vision Transformer (ViT) has shown impressive performance on high-level and low-level vision tasks. In this paper, we propose a new ViT architecture, named Hybrid Local-Global Vision Transformer (HyLoG-ViT), for single image dehazing. T he HyLoG-ViT block consists of two paths, the local ViT path and the global ViT path, which are used to capture local and global dependencies. The hybrid features are fused via convolution layers. As a result, the HyLoG-ViT reduces the computational complexity and introduces locality in the networks. Then, the HyLoG-ViT blocks are incorporated within our dehazing networks, which jointly learn the intrinsic image decomposition and image dehazing. Specifically, the network consists of one shared encoder and three decoders for reflectance prediction, shading prediction, and haze-free image generation. The tasks of reflectance and shading prediction can produce meaningful intermediate features that can serve as complementary features for haze-free image generation. To effectively aggregate the complementary features, we propose a complementary features selection module (CFSM) to select the useful ones for image dehazing. Extensive experiments on homogeneous, non-homogeneous, and nighttime dehazing tasks reveal that our proposed Transformer-based dehazing network can achieve comparable or even better performance than CNNs-based dehazing models.
We present a neural architecture search (NAS) technique to enhance the performance of unsupervised image de-noising, in-painting and super-resolution under the recently proposed Deep Image Prior (DIP). We show that evolutionary search can automatical ly optimize the encoder-decoder (E-D) structure and meta-parameters of the DIP network, which serves as a content-specific prior to regularize these single image restoration tasks. Our binary representation encodes the design space for an asymmetric E-D network that typically converges to yield a content-specific DIP within 10-20 generations using a population size of 500. The optimized architectures consistently improve upon the visual quality of classical DIP for a diverse range of photographic and artistic content.
Image retrieval is the problem of searching an image database for items that are similar to a query image. To address this task, two main types of image representations have been studied: global and local image features. In this work, our key contrib ution is to unify global and local features into a single deep model, enabling accurate retrieval with efficient feature extraction. We refer to the new model as DELG, standing for DEep Local and Global features. We leverage lessons from recent feature learning work and propose a model that combines generalized mean pooling for global features and attentive selection for local features. The entire network can be learned end-to-end by carefully balancing the gradient flow between two heads -- requiring only image-level labels. We also introduce an autoencoder-based dimensionality reduction technique for local features, which is integrated into the model, improving training efficiency and matching performance. Comprehensive experiments show that our model achieves state-of-the-art image retrieval on the Revisited Oxford and Paris datasets, and state-of-the-art single-model instance-level recognition on the Google Landmarks dataset v2. Code and models are available at https://github.com/tensorflow/models/tree/master/research/delf .
Non-Local (NL) blocks have been widely studied in various vision tasks. However, it has been rarely explored to embed the NL blocks in mobile neural networks, mainly due to the following challenges: 1) NL blocks generally have heavy computation cost which makes it difficult to be applied in applications where computational resources are limited, and 2) it is an open problem to discover an optimal configuration to embed NL blocks into mobile neural networks. We propose AutoNL to overcome the above two obstacles. Firstly, we propose a Lightweight Non-Local (LightNL) block by squeezing the transformation operations and incorporating compact features. With the novel design choices, the proposed LightNL block is 400x computationally cheaper} than its conventional counterpart without sacrificing the performance. Secondly, by relaxing the structure of the LightNL block to be differentiable during training, we propose an efficient neural architecture search algorithm to learn an optimal configuration of LightNL blocks in an end-to-end manner. Notably, using only 32 GPU hours, the searched AutoNL model achieves 77.7% top-1 accuracy on ImageNet under a typical mobile setting (350M FLOPs), significantly outperforming previous mobile models including MobileNetV2 (+5.7%), FBNet (+2.8%) and MnasNet (+2.1%). Code and models are available at https://github.com/LiYingwei/AutoNL.
101 - Xiu Su , Shan You , Jiyang Xie 2021
Recently, transformers have shown great superiority in solving computer vision tasks by modeling images as a sequence of manually-split patches with self-attention mechanism. However, current architectures of vision transformers (ViTs) are simply inh erited from natural language processing (NLP) tasks and have not been sufficiently investigated and optimized. In this paper, we make a further step by examining the intrinsic structure of transformers for vision tasks and propose an architecture search method, dubbed ViTAS, to search for the optimal architecture with similar hardware budgets. Concretely, we design a new effective yet efficient weight sharing paradigm for ViTs, such that architectures with different token embedding, sequence size, number of heads, width, and depth can be derived from a single super-transformer. Moreover, to cater for the variance of distinct architectures, we introduce textit{private} class token and self-attention maps in the super-transformer. In addition, to adapt the searching for different budgets, we propose to search the sampling probability of identity operation. Experimental results show that our ViTAS attains excellent results compared to existing pure transformer architectures. For example, with $1.3$G FLOPs budget, our searched architecture achieves $74.7%$ top-$1$ accuracy on ImageNet and is $2.5%$ superior than the current baseline ViT architecture. Code is available at url{https://github.com/xiusu/ViTAS}.
التعليقات
جاري جلب التعليقات جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
mircosoft-partner

هل ترغب بارسال اشعارات عن اخر التحديثات في شمرا-اكاديميا