To generate new images for a given category, most deep generative models require abundant training images from this category, which are often too expensive to acquire. To achieve the goal of generation based on only a few images, we propose matching-based Generative Adversarial Network (GAN) for few-shot generation, which includes a matching generator and a matching discriminator. Matching generator can match random vectors with a few conditional images from the same category and generate new images for this category based on the fused features. The matching discriminator extends conventional GAN discriminator by matching the feature of generated image with the fused feature of conditional images. Extensive experiments on three datasets demonstrate the effectiveness of our proposed method.
Static image action recognition, which aims to recognize action based on a single image, usually relies on expensive human labeling effort such as adequate labeled action images and large-scale labeled image dataset. In contrast, abundant unlabeled videos can be economically obtained. Therefore, several works have explored using unlabeled videos to facilitate image action recognition, which can be categorized into the following two groups: (a) enhance visual representations of action images with a designed proxy task on unlabeled videos, which falls into the scope of self-supervised learning; (b) generate auxiliary representations for action images with the generator learned from unlabeled videos. In this paper, we integrate the above two strategies in a unified framework, which consists of Visual Representation Enhancement (VRE) module and Motion Representation Augmentation (MRA) module. Specifically, the VRE module includes a proxy task which imposes pseudo motion label constraint and temporal coherence constraint on unlabeled videos, while the MRA module could predict the motion information of a static action image by exploiting unlabeled videos. We demonstrate the superiority of our framework based on four benchmark human action datasets with limited labeled data.
The goal of Sketch-Based Image Retrieval (SBIR) is using free-hand sketches to retrieve images of the same category from a natural image gallery. However, SBIR requires all categories to be seen during training, which cannot be guaranteed in real-world applications. So we investigate more challenging Zero-Shot SBIR (ZS-SBIR), in which test categories do not appear in the training stage. Traditional SBIR methods are prone to be category-based retrieval and cannot generalize well from seen categories to unseen ones. In contrast, we disentangle image features into structure features and appearance features to facilitate structure-based retrieval. To assist feature disentanglement and take full advantage of disentangled information, we propose a Bi-directional Domain Translation (BDT) framework for ZS-SBIR, in which the image domain and sketch domain can be translated to each other through disentangled structure and appearance features. Finally, we perform retrieval in both structure feature space and image feature space. Extensive experiments demonstrate that our proposed approach remarkably outperforms state-of-the-art approaches by about 8% on the Sketchy dataset and over 5% on the TU-Berlin dataset.
Image composition is an important operation in image processing, but the inconsistency between foreground and background significantly degrades the quality of composite image. Image harmonization, aiming to make the foreground compatible with the background, is a promising yet challenging task. However, the lack of high-quality publicly available dataset for image harmonization greatly hinders the development of image harmonization techniques. In this work, we contribute an image harmonization dataset by generating synthesized composite images based on COCO (resp., Adobe5k, Flickr, day2night) dataset, leading to our HCOCO (resp., HAdobe5k, HFlickr, Hday2night) sub-dataset. Moreover, we propose a new deep image harmonization method with a novel domain verification discriminator, enlightened by the following insight. Specifically, incompatible foreground and background belong to two different domains, so we need to translate the domain of foreground to the same domain as background. Our proposed domain verification discriminator can play such a role by pulling close the domains of foreground and background. Extensive experiments on our constructed dataset demonstrate the effectiveness of our proposed method. Our dataset is released in https://github.com/bcmi/Image_Harmonization_Datasets.
Activity image-to-video retrieval task aims to retrieve videos containing the similar activity as the query image, which is a challenging task because videos generally have many background segments irrelevant to the activity. In this paper, we utilize R-C3D model to represent a video by a bag of activity proposals, which can filter out background segments to some extent. However, there are still noisy proposals in each bag. Thus, we propose an Activity Proposal-based Image-to-Video Retrieval (APIVR) approach, which incorporates multi-instance learning into cross-modal retrieval framework to address the proposal noise issue. Specifically, we propose a Graph Multi-Instance Learning (GMIL) module with graph convolutional layer, and integrate this module with classification loss, adversarial loss, and triplet loss in our cross-modal retrieval framework. Moreover, we propose geometry-aware triplet loss based on point-to-subspace distance to preserve the structural information of activity proposals. Extensive experiments on three widely-used datasets verify the effectiveness of our approach.
Aesthetic image cropping is a practical but challenging task which aims at finding the best crops with the highest aesthetic quality in an image. Recently, many deep learning methods have been proposed to address this problem, but they did not reveal the intrinsic mechanism of aesthetic evaluation. In this paper, we propose an interpretable image cropping model to unveil the mystery. For each image, we use a fully convolutional network to produce an aesthetic score map, which is shared among all candidate crops during crop-level aesthetic evaluation. Then, we require the aesthetic score map to be both composition-aware and saliency-aware. In particular, the same region is assigned with different aesthetic scores based on its relative positions in different crops. Moreover, a visually salient region is supposed to have more sensitive aesthetic scores so that our network can learn to place salient objects at more proper positions. Such an aesthetic score map can be used to localize aesthetically important regions in an image, which sheds light on the composition rules learned by our model. We show the competitive performance of our model in the image cropping task on several benchmark datasets, and also demonstrate its generality in real-world applications.
Image composition is an important operation in image processing, but the inconsistency between foreground and background significantly degrades the quality of composite image. Image harmonization, which aims to make the foreground compatible with the background, is a promising yet challenging task. However, the lack of high-quality public dataset for image harmonization, which significantly hinders the development of image harmonization techniques. Therefore, we create synthesized composite images based on existing COCO (resp., Adobe5k, day2night) dataset, leading to our HCOCO (resp., HAdobe5k, Hday2night) dataset. To enrich the diversity of our datasets, we also generate synthesized composite images based on our collected Flick images, leading to our HFlickr dataset. All four datasets are released in https://github.com/bcmi/Image_Harmonization_Datasets.
Learning from web data has attracted lots of research interest in recent years. However, crawled web images usually have two types of noises, label noise and background noise, which induce extra difficulties in utilizing them effectively. Most existing methods either rely on human supervision or ignore the background noise. In this paper, we propose the novel ProtoNet, which is capable of handling these two types of noises together, without the supervision of clean images in the training stage. Particularly, we use a memory module to identify the representative and discriminative prototypes for each category. Then, we remove noisy images and noisy region proposals from the web dataset with the aid of the memory module. Our approach is efficient and can be easily integrated into arbitrary CNN model. Extensive experiments on four benchmark datasets demonstrate the effectiveness of our method.
Semantic segmentation has achieved significant progress but is still challenging due to the complex scene, object occlusion, and so on. Some research works have attempted to use extra information such as depth information to help RGB based semantic segmentation. However, extra information is usually unavailable for the test images. Inspired by learning using privileged information, in this paper, we only leverage the depth information of training images as privileged information in the training stage. Specifically, we rely on depth information to identify the hard pixels which are difficult to classify, by using our proposed Depth Prediction Error (DPE) and Depth-dependent Segmentation Error (DSE). By paying more attention to the identified hard pixels, our approach achieves the state-of-the-art results on two benchmark datasets and even outperforms the methods which use depth information of test images.