Search

Photo-realistic image synthesis is an attracting idea for person re-identification (ReID) and data augmentation on human pose estimation. However, existing advances manipulating human image synthesis lack texture details for varying poses or appearances. This paper presents a person image synthesis Siamese generative adversarial network (PS2GAN), which re-synthesizes person image by changing the pose of that person to a given pose, modeled in a Siamese structure with image generative network and pair conditional discriminative networks in single-branch. For pose transfer, the proposed PS2GAN adopts Siamese structure consisting of two image generative networks and a novel contrastive-pose loss regularizing the generation process. Additionally, a nearest-neighbor loss computes the difference between fake and real images to make high-level information closer. Furthermore, the proposed PS2GAN is competitive to the state-of-the-art performance on Market-1501 and DeepFashion datasets via qualitatively and quantitatively comparing with prior works, and synthetic images of the PS2GAN can alleviate data insufficiency for person ReID.

In recent years, convolutional neural networks (CNN) have been extensively used for generic object detection due to their powerful feature extraction capabilities. This has hence motivated researchers to adopt this technology in the field of remote sensing. However, remote sensing images can contain large amounts of noise, have complex backgrounds, include small dense objects as well as being susceptible to weather and light intensity variations. Moreover, from different shooting angles, objects can either have different shapes or be obscured by structures such as buildings and trees. Due to these, effective features extraction for proper representation is still very challenging from remote sensing images. This paper therefore proposes a novel remote sensing image object detection approach applying a fusion-based feature reinforcement component (FB-FRC) to improve the discrimination between object feature. Specifically, two fusion strategies are proposed: (i) a hard fusion strategy through artificially-set rules, and (ii) a soft fusion strategy by learning the fusion parameters. Experiments carried out on four widely used remote sensing datasets (NWPU VHR-10, VisDrone2018, DOTA and RSOD) have shown promising results where the proposed approach manages to outperform several state-of-the-art methods.

The remote sensing data is difficult to collect and lack of diversity, which extremely limits the performance of object detection on remote sensing images. In this paper, a multi-branch conditional generative adversarial network (MCGAN) is proposed to augment data for object detection in optical remote sensing images, which is the first GANs-based data augmentation framework proposed for this topic. We use MCGAN to generate the diverse objects based on the existing remote sensing datasets. The multi-branch dilated convolution and the classification branch are adopted into MCGAN to help the generator to generate the diverse and high-quality images. Meanwhile, an adaptive samples selection strategy based on the Faster R-CNN is proposed to select the samples for data augmentation from the objects generated by MCGAN, which can ensure the quality of new augmented training sets and improve the diversity of samples. Experiments based on NWPU VHR-10 and DOTA show that the objects generated by MCGAN have the higher quality compared with the objects generated by WGAN and LSGAN. And the mean average precision detected by the state-of-the-art object detection models used in the experiments has the satisfactory improvement after the MCGAN based data augmentation, which indicates that data augmentation by MCGAN can effectively improve the accuracy of remote sensing images object detection.

Person image synthesis based on shape and appearance using deep generative models opens the door in mickle applications, such as person re-identification (ReID) and movie industry. The methods of image synthesis are driven by producing the image of an object directly, which fail to recover spatial deformations when images are generated. In this paper, we present a conditional variational generative adversarial network (CVGAN) to synthesize desired images guided by target shape by modeling the inherent interplay between shape and appearance. Firstly, the shape and appearance of the given images are disentangled by adopting variational inference, which enables us to generate person images with arbitrary shapes. Secondly, to preserve the details and generate photo-realistic images, the Kullback–Leibler (KL) loss is adopted to reduce the gap between the condition image and generated image. Thirdly, to prevent partly gradient vanishing problem for training our framework stably, we propose combined general learning method, where the discriminative network leverages least squares loss. In addition, we experiment on COCO, DeepFashion and Market-1501 datasets, and results demonstrate that VGAN significantly improves the synthesis of images on discriminability, diversity and quality over the existing methods.

Deep convolutional neural networks have been widely used in many areas. Generally, a vast amount of data are required for deep neural networks training, since they have a large number of parameters. This paper devotes to develop a many-objective convolutional neural network (MaO-CNN) model, which can obtain better classification performance than a single-objective one without sufficient training data. The main contributions of this paper are listed as follows: firstly, we propose many-class detection error trade-off (MaDET) and develop a MaO-CNN model in MaDET space; secondly, a hybrid framework of many-objective evolutionary algorithm is proposed for MaO-CNN model training; thirdly, a encoding method is designed for parameters encoding and MaO-CNN evolving. Experimental results based on well-known MNIST and SVHN datasets show that the new proposed model can obtain better results than a conventional one with the same amount of training data.