Search

Scene image is usually composed of background information and foreground objects.Convolutional neural network (CNN) used for scene recognition task usually needs to recognize the category of scene according to the characteristics of key objects in the scene, or even combined with the position relationship between objects.Aiming at the problem that the key target features of small size in the scene image gradually disappear with the deepening of the network level, which leads to scene recognition errors, a scene recognition method based on multi-level feature fusion and attention module is proposed.Firstly, the feature extraction part of the deep neural network ResNet-18 is divided into five branches, and then the multi-level features of the output of the five branches are fused, and the fused features are used for scene recognition and classification to make up for the lost target information.Secondly, an improved attention module is added to the network to achieve the purpose of focusing on learning the key targets in the scene image, so as to improve the recognition effect further.Experimental results on several scene datasets show that the recognition accuracy of the proposed method on MIT-67, SUN-397 and UIUC-Sports scene datasets reaches 88.2%, 79.9% and 97.7% respectively, which is higher than the current mainstream scene recognition methods.

Consistency-based semi-supervised learning methods typically use simple data augmentation methods to achieve consistent predictions for both original inputs and perturbed inputs.The effectiveness of this approach is difficult to be guaranteed when the proportion of labeled data is relatively low.Extending some advanced data augmentation method in supervised learning to be used in a semi-supervised learning setting is one of the ideas to solve this problem.Based on the consistency-based semi-supervised learning method MixMatch,a semi-supervised learning method AutoMixMatch based on automated mixed sample data augmentation techniques is proposed,which uses a modified automatic data augmentation technique in the data augmentation phase,and a mixed-sample algorithm is proposed to enhance the utilization of unlabeled samples in the sample mixing phase.The performance of the proposed method is evaluated through image classification experiments.In image classification benchmark datasets,the proposed method outperforms several mainstream semi-supervised classification methods in three labeled sample proportions,which validates the effectiveness of the method.In addition,the proposed method performs better with a very low proportion of labeled data to the training data (only 0.05%),and the classification error rate of the proposed method on the SVHN dataset is 30.17% lower than that of MixMatch.

The purpose of material recognition is to identify the main objects and their material categories in natural material images.Aiming at the problem of low recognition accuracy caused by the lack of data in material image data sets and the difficulty of manually labeling local texture regions,a material recognition method based on attention mechanism and deep convolutional neural network is proposed.The core of the method is material recognition deep convolutional neural network (MaterialNet).MaterialNet uses the deep residual network to extract the features of the image,and introduces the attention mechanism by the proposed cascaded atrous spatial pyramid pooling method,so that the network can adaptively focus on the key areas containing texture features through end-to-end training,so as to effectively identify the local texture features of materials.Based on the FMD material datasets,the experimental results show that the overall identification accuracy of MaterialNet is 82.3%,which is 7.2% and 4.5% higher than the current mainstream B-CNN and CNN+FV material identification methods,respectively.The recognition accuracy of MaterialNet is high for a variety of materials,and it has the advantages of less parameters and less calculation.

In order to solve the problems of high cost of obtaining the top plan view of the main outline of the building in the 3D modeling scenario,low segmentation accuracy of the aerial photography building,interference on the roof of the building,etc.,a method of accurately segmenting the aerial photography building based on deep residual network is proposed,in which the positions of five points are expressed as heat maps as additional input channels of the network,and good segmentation effect is achieved in the task of accurately segmenting the aerial photography building.Experimental results show that the proposedmethod has higher segmentation accuracy and segmentation efficiency than the traditional semi-automatic segmentation method Grabcut.It has better robustness and anti-interference than DEXTR method.This method can provide high-precision top-view contour map and top-view picture of buildings for 3D reconstruction of buildings,and can also be used in the production process of aerial photography building data sets as an accurate and effective mask annotation tool or semi-automatic contour annotation tool to improve the annotation efficiency of datasets.

C6H25Br3N2O3P2, monoclinic, P21/c (no. 14), a = 8.9705(10) Å, b = 25.494(3) Å, c = 10.2372(12) Å, β = 112.888(2)°, V = 2156.8(49) Å3, Z = 4, Rgt(F) = 0.0467, wRref(F2) = 0.1296, T = 296(2) K.

C18H20ClN3O, monoclinic, P21/c (no. 14), a = 10.0099(3) Å, b = 16.2337(6) Å, c = 21.0714(7) Å, β = 91.324(3)°, V = 3423.1(2) Å3, Z = 8, Rgt(F) = 0.0381, wRref(F2) = 0.1065, T = 291(2) K.

C32H28N2O2, monoclinic, P21/n (no. 14), a = 10.1291(10) Å, b = 25.7313(3) Å, c = 10.2395(10) Å, β = 106.521(1)°, V = 2558.6(4) Å3, Z = 4, Rgt(F) = 0.0402, wRref(F2) = 0.1045, T = 296(2) K.

C24H24N2O2, monoclinic, P21/c (no. 14), a = 11.7857(4) Å, b = 10.6233(4) Å, c = 8.0268(3) Å, β = 102.104(2)°, V = 982.64(6) Å3, Z = 2, Rgt(F) = 0.0320, wRref(F2) = 0.0827, T = 293(2) K.

C10H9N3O3, monoclinic, P21/c (no. 14), a = 11.290(2) Å, b = 7.339(1) Å, c = 12.238(2) Å, β = 101.325(3)°, V = 994.3 Å3, Z = 4, Rgt(F) = 0.0408, wRref(F2) = 0.1209, T = 296 K.

C6H25Br3N2O3P2, monoclinic, P21/ c (no. 14), a = 8.9705(10) Å, b = 25.494(3) Å, c = 10.2372(12) Å, β = 112.888(2)°, V = 2156.8(49) Å3, Z = 4, Rgt( F) = 0.0467, wRref( F2) = 0.1296, T = 296(2) K. [ABSTRACT FROM AUTHOR]

C6H25Br3N2O3P2, monoclinic, P21/ c (no. 14), a = 8.9705(10) Å, b = 25.494(3) Å, c = 10.2372(12) Å, β = 112.888(2)°, V = 2156.8(49) Å3, Z = 4, Rgt( F) = 0.0467, wRref( F2) = 0.1296, T = 296(2) K. [ABSTRACT FROM AUTHOR]

C32H28N2O2, monoclinic, P21/ n (no. 14), a = 10.1291(10) Å, b = 25.7313(3) Å, c = 10.2395(10) Å, β = 106.521(1)°, V = 2558.6(4) Å3, Z = 4, Rgt (F) = 0.0402, wRref (F2 ) = 0.1045, T = 296(2) K. [ABSTRACT FROM AUTHOR]

Copyright of Zeitschrift für Kristallographie / New Crystal Structures is the property of De Gruyter and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

C12H12Cl2N2Zn, monoclinic, P21/ c (no. 14), a = 8.615(2) Å, b = 10.589(2) Å, c = 16.525(2) Å, β = 116.598(6)°, V = 1347.9 Å3, Z = 4, Rgt (F) = 0.0247, wRref (F2 ) = 0.0697, T = 298 K. [ABSTRACT FROM AUTHOR]

C10H9N3O3, monoclinic, P21/c (no. 14), a = 11.290(2) Å, b = 7.339(1) Å, c = 12.238(2) Å, β = 101.325(3)°, V = 994.3 Å3, Z = 4, Rgt(F) = 0.0408, wRref(F2) = 0.1209, T = 296 K. [ABSTRACT FROM AUTHOR]

C32H28N2O2, monoclinic, P21/ n (no. 14), a = 10.1291(10) Å, b = 25.7313(3) Å, c = 10.2395(10) Å, β = 106.521(1)°, V = 2558.6(4) Å3, Z = 4, Rgt (F) = 0.0402, wRref (F2 ) = 0.1045, T = 296(2) K. [ABSTRACT FROM AUTHOR]

The title compound, C(15)H(9)ClN(2)O(2), adopts an E configuration about the C=N double bond. The mean plane of the isoindoline ring system [maximum deviation = 0.011 (2) Å] is inclined to the chloro-benzene ring by 22.62 (8)°. In the crystal, mol-ecules are connected by C-H⋯O hydrogen bonds, forming chains that propagate along [010].

The title compound, C(19)H(12)N(2)O(3), has two independent mol-ecules (A and B) in the asymmetric unit. There is an intra-molecular O-H⋯N hydrogen bond in each mol-ecule. The mean planes of the naphthalene [maximum deviations = 0.024 (3) and 0.030 (3) Å in A and B, respectively] and the isoindoline units [maximum deviations 0.009 (3) and 0.008 (3) Å in A and B, respectively] are almostly coplanar, with dihedral angles of 4.25 (9) ad 3.84 (9)° in mol-ecules A and B, respectively. The two independent mol-ecules are connected by π-π inter-actions [centroid-centroid distances 3.5527 (19) and 3.5627 (19) Å]. In the crystal, the A+B pairs are further connected via π-π inter-actions [centroid-centroid distances = 3.693 (2)-3.831 (2) Å], leading to the formation of columns propagating along the a-axis direction. The columns are linked via C-H⋯O inter-actions, leading to the formation of a three-dimensional network.

In the title compound, C(16)H(16)N(4)O(6), the planes of the isoindole and dinitro-benzene groups make a dihedral angle between of 84.15 (8)°. The N atom of the isoindole group is displaced by 0.2937 (3) Å from the plane through the remaining atoms. An intra-molecular N-H⋯O inter-action occurs. In the crystal, inversion dimers linked by pairs of N-H⋯O hydrogen bonds occur.

The title compound, C(19)H(21)N(3)O(4), crystallizes with two independent mol-ecules in the asymmetric unit. In both mol-ecules, there is an intra-molecular O-H⋯N hydrogen bond, which correlates with the fact that each mol-ecule adopts an E configuration with respect to the C=N bond. In the crystal, there are C-H⋯O and C-H⋯π inter-actions present.

The title compound, C(13)H(12)N(2)O(3), has an E configuration with respect to the C=N bond: the conformation is stabilized by an intramolecular O-H⋯N hydrogen bond. In the crystal, an N-H⋯O interaction links the molecules into a C(4) chain along [100].

The title compound, C(14)H(11)N(3)O(3), adopts an E or trans configuration with respect to the C=N bond. In the mol-ecule there is an intra-molecular O-H⋯N hydrogen bond involving the hy-droxy substituent at the 2-positon of the naphthalene ring and the adjacent methyl-ene-amino N atom. The mol-ecule is roughly planar, the dihedral angle between the naphthalene and imidazolidine-2,4-dione mean planes being 8.4 (1)°. In the crystal, pairs of N-H⋯O hydrogen bonds link mol-ecules into inversion dimers. These dimers are futher linked via C-H⋯O inter-actions, forming a three-dimensional network.

The title mol-ecule, C(14)H(10)N(4)O(6), crystallizes with one half-mol-ecule in the asymmetric unit; the mid-point of the N-N bond lies on an inversion centre. The nitro and amide groups are twisted with respect to the benzene ring, making dihedral angles of 14.6 (5) and 31.1 (5)°, respectively. In the crystal structure, mol-ecules are linked through N-H⋯O hydrogen bonding between the imino and carbonyl groups.