Arbitrary-shaped text detection is a challenging task due to the complex geometric layouts of texts such as large aspect ratios, various scales, random rotations and curve shapes. Most state-of-the-art methods solve this problem from bottom-up perspectives, seeking to model a text instance of complex geometric layouts with simple local units (e.g., local boxes or pixels) and generate detections with heuristic post-processings. In this work, we propose an arbitrary-shaped text detection method, namely TextRay, which conducts top-down contour-based geometric modeling and geometric parameter learning within a single-shot anchor-free framework. The geometric modeling is carried out under polar system with a bidirectional mapping scheme between shape space and parameter space, encoding complex geometric layouts into unified representations. For effective learning of the representations, we design a central-weighted training strategy and a content loss which builds propagation paths between geometric encodings and visual content. TextRay outputs simple polygon detections at one pass with only one NMS post-processing. Experiments on several benchmark datasets demonstrate the effectiveness of the proposed approach. The code is available at https://github.com/LianaWang/TextRay.
Visual Storytelling~(VIST) is a task to tell a narrative story about a certain topic according to the given photo stream. The existing studies focus on designing complex models, which rely on a huge amount of human-annotated data. However, the annotation of VIST is extremely costly and many topics cannot be covered in the training dataset due to the long-tail topic distribution. In this paper, we focus on enhancing the generalization ability of the VIST model by considering the few-shot setting. Inspired by the way humans tell a story, we propose a topic adaptive storyteller to model the ability of inter-topic generalization. In practice, we apply the gradient-based meta-learning algorithm on multi-modal seq2seq models to endow the model the ability to adapt quickly from topic to topic. Besides, We further propose a prototype encoding structure to model the ability of intra-topic derivation. Specifically, we encode and restore the few training story text to serve as a reference to guide the generation at inference time. Experimental results show that topic adaptation and prototype encoding structure mutually bring benefit to the few-shot model on BLEU and METEOR metric. The further case study shows that the stories generated after few-shot adaptation are more relative and expressive.
With the memory-resource-limited constraints, class-incremental learning (CIL) usually suffers from the "catastrophic forgetting" problem when updating the joint classification model on the arrival of newly added classes. To cope with the forgetting problem, many CIL methods transfer the knowledge of old classes by preserving some exemplar samples into the size-constrained memory buffer. To utilize the memory buffer more efficiently, we propose to keep more auxiliary low-fidelity exemplar samples rather than the original real high-fidelity exemplar samples. Such memory-efficient exemplar preserving scheme make the old-class knowledge transfer more effective. However, the low-fidelity exemplar samples are often distributed in a different domain away from that of the original exemplar samples, that is, a domain shift. To alleviate this problem, we propose a duplet learning scheme that seeks to construct domain-compatible feature extractors and classifiers, which greatly narrows down the above domain gap. As a result, these low-fidelity auxiliary exemplar samples have the ability to moderately replace the original exemplar samples with a lower memory cost. In addition, we present a robust classifier adaptation scheme, which further refines the biased classifier (learned with the samples containing distillation label knowledge about old classes) with the help of the samples of pure true class labels. Experimental results demonstrate the effectiveness of this work against the state-of-the-art approaches. We will release the code, baselines, and training statistics for all models to facilitate future research.
Learning to generate natural scenes has always been a daunting task in computer vision. This is even more laborious when generating images with very different views. When the views are very different, the view fields have little overlap or objects are occluded, leading the task very challenging. In this paper, we propose to use Generative Adversarial Networks(GANs) based on a deformable convolution and attention mechanism to solve the problem of cross-view image synthesis (see Fig.1). It is difficult to understand and transform scenes appearance and semantic information from another view, thus we use deformed convolution in the U-net network to improve the network's ability to extract features of objects at different scales. Moreover, to better learn the correspondence between images from different views, we apply an attention mechanism to refine the intermediate feature map thus generating more realistic images. A large number of experiments on different size images on the Dayton dataset[1] show that our model can produce better results than state-of-the-art methods.
In real applications, object detectors based on deep networks still face challenges of the large domain gap between the labeled training data and unlabeled testing data. To reduce the gap, recent techniques are proposed by aligning the image/instance-level features between source and unlabeled target domains. However, these methods suffer from the suboptimal problem mainly because of ignoring the category information of object instances. To tackle this issue, we develop a fine-grained domain alignment approach with a well-designed domain classifier bank that achieves the instance-level alignment respecting to their categories. Specifically, we first employ the mean teacher paradigm to generate pseudo labels for unlabeled samples. Then we implement the class-level domain classifiers and group them together, called domain classifier bank, in which each domain classifier is responsible for aligning features of a specific class. We assemble the bare object detector with the proposed fine-grained domain alignment mechanism as the adaptive detector, and optimize it with a developed crossed adaptive weighting mechanism. Extensive experiments on three popular transferring benchmarks demonstrate the effectiveness of our method and achieve the new remarkable state-of-the-arts.
Federated learning enables collaboratively training machine learning models on decentralized data. The three types of heterogeneous natures that is data, model, and objective bring about unique challenges to the canonical federated learning algorithm (FederatedAveraging), where one shared model is produced by and for all clients. First, due to the Non-IIDness of data, the global shared model may perform worse than local models that solely trained on their private data; Second, clients may need to design their own model because of different communication and computing abilities of devices, which is also private property that should be protected; Third, the objective of achieving consensus throughout the training process will compromise the personalities of clients. In this work, we present a novel federated learning paradigm, named Federated Mutual Leaning (FML), dealing with the three heterogeneities. FML allows clients designing their customized models and training independently, thus the Non-IIDness of data is no longer a bug but a feature that clients can be personally served better. Local customized models can benefit from collaboratively training without compromising personalities. Global model does not have to be an out-of-the-box (OOTB) product but a meta-learner which requires local adaptation for new participants. The experiments show that FML can achieve better performance, robustness and communication efficiency than alternatives.
In e-commerce, consumer-generated videos, which in general deliver consumers' individual preferences for the different aspects of certain products, are massive in volume. To recommend these videos to potential consumers more effectively, diverse and catchy video titles are critical. However, consumer-generated videos seldom accompany appropriate titles. To bridge this gap, we integrate comprehensive sources of information, including the content of consumer-generated videos, the narrative comment sentences supplied by consumers, and the product attributes, in an end-to-end modeling framework. Although automatic video titling is very useful and demanding, it is much less addressed than video captioning. The latter focuses on generating sentences that describe videos as a whole while our task requires the product-aware multi-grained video analysis. To tackle this issue, the proposed method consists of two processes, i.e., granular-level interaction modeling and abstraction-level story-line summarization. Specifically, the granular-level interaction modeling first utilizes temporal-spatial landmark cues, descriptive words, and abstractive attributes to builds three individual graphs and recognizes the intra-actions in each graph through Graph Neural Networks (GNN). Then the global-local aggregation module is proposed to model inter-actions across graphs and aggregate heterogeneous graphs into a holistic graph representation. The abstraction-level story-line summarization further considers both frame-level video features and the holistic graph to utilize the interactions between products and backgrounds, and generate the story-line topic of the video. We collect a large-scale dataset accordingly from real-world data in Taobao, a world-leading e-commerce platform, and will make the desensitized version publicly available to nourish further development of the research community...
Recently, remarkable progress has been made in weakly supervised object localization (WSOL) to promote object localization maps. The common practice of evaluating these maps applies an indirect and coarse way, i.e., obtaining tight bounding boxes which can cover high-activation regions and calculating intersection-over-union (IoU) scores between the predicted and ground-truth boxes. This measurement can evaluate the ability of localization maps to some extent, but we argue that the maps should be measured directly and delicately, i.e., comparing the maps with the ground-truth object masks pixel-wisely. To fulfill the direct evaluation, we annotate pixel-level object masks on the ILSVRC validation set. We propose to use IoU-Threshold curves for evaluating the real quality of localization maps. Beyond the amended evaluation metric and annotated object masks, this work also introduces a novel self-enhancement method to harvest accurate object localization maps and object boundaries with only category labels as supervision. We propose a two-stage approach to generate the localization maps by simply comparing the similarity of point-wise features between the high-activation and the rest pixels. Based on the predicted localization maps, we explore to estimate object boundaries on a very large dataset. A hard-negative suppression loss is proposed for obtaining fine boundaries. We conduct extensive experiments on the ILSVRC and CUB benchmarks. In particular, the proposed Self-Enhancement Maps achieve the state-of-the-art localization accuracy of 54.88% on ILSVRC. The code and the annotated masks are released at https://github.com/xiaomengyc/SEM.
One fundamental problem in the learning treatment effect from observational data is confounder identification and balancing. Most of the previous methods realized confounder balancing by treating all observed variables as confounders, ignoring the identification of confounders and non-confounders. In general, not all the observed variables are confounders which are the common causes of both the treatment and the outcome, some variables only contribute to the treatment and some contribute to the outcome. Balancing those non-confounders would generate additional bias for treatment effect estimation. By modeling the different relations among variables, treatment and outcome, we propose a synergistic learning framework to 1) identify and balance confounders by learning decomposed representation of confounders and non-confounders, and simultaneously 2) estimate the treatment effect in observational studies via counterfactual inference. Our empirical results demonstrate that the proposed method can precisely identify and balance confounders, while the estimation of the treatment effect performs better than the state-of-the-art methods on both synthetic and real-world datasets.