arXiv subjects

Jun Xiao

Publications and source records attributed to Jun Xiao.

At least 253 records · Page 14Linked to original sources

arXiv2020

Counterfactual Samples Synthesizing for Robust Visual Question Answering

Despite Visual Question Answering (VQA) has realized impressive progress over the last few years, today's VQA models tend to capture superficial linguistic correlations in the train set and fail to generalize to the test set with different QA distributions. To reduce the language biases, several recent works introduce an auxiliary question-only model to regularize the training of targeted VQA model, and achieve dominating performance on VQA-CP. However, since the complexity of design, current methods are unable to equip the ensemble-based models with two indispensable characteristics of an ideal VQA model: 1) visual-explainable: the model should rely on the right visual regions when making decisions. 2) question-sensitive: the model should be sensitive to the linguistic variations in question. To this end, we propose a model-agnostic Counterfactual Samples Synthesizing (CSS) training scheme. The CSS generates numerous counterfactual training samples by masking critical objects in images or words in questions, and assigning different ground-truth answers. After training with the complementary samples (ie, the original and generated samples), the VQA models are forced to focus on all critical objects and words, which significantly improves both visual-explainable and question-sensitive abilities. In return, the performance of these models is further boosted. Extensive ablations have shown the effectiveness of CSS. Particularly, by building on top of the model LMH, we achieve a record-breaking performance of 58.95% on VQA-CP v2, with 6.5% gains.

cs.CV↗

arXiv2020

Evaluation Framework For Large-scale Federated Learning

Federated learning is proposed as a machine learning setting to enable distributed edge devices, such as mobile phones, to collaboratively learn a shared prediction model while keeping all the training data on device, which can not only take full advantage of data distributed across millions of nodes to train a good model but also protect data privacy. However, learning in scenario above poses new challenges. In fact, data across a massive number of unreliable devices is likely to be non-IID (identically and independently distributed), which may make the performance of models trained by federated learning unstable. In this paper, we introduce a framework designed for large-scale federated learning which consists of approaches to generating dataset and modular evaluation framework. Firstly, we construct a suite of open-source non-IID datasets by providing three respects including covariate shift, prior probability shift, and concept shift, which are grounded in real-world assumptions. In addition, we design several rigorous evaluation metrics including the number of network nodes, the size of datasets, the number of communication rounds and communication resources etc. Finally, we present an open-source benchmark for large-scale federated learning research.

cs.LG↗

arXiv2020

Reinforcement-Learning based Portfolio Management with Augmented Asset Movement Prediction States

Portfolio management (PM) is a fundamental financial planning task that aims to achieve investment goals such as maximal profits or minimal risks. Its decision process involves continuous derivation of valuable information from various data sources and sequential decision optimization, which is a prospective research direction for reinforcement learning (RL). In this paper, we propose SARL, a novel State-Augmented RL framework for PM. Our framework aims to address two unique challenges in financial PM: (1) data heterogeneity -- the collected information for each asset is usually diverse, noisy and imbalanced (e.g., news articles); and (2) environment uncertainty -- the financial market is versatile and non-stationary. To incorporate heterogeneous data and enhance robustness against environment uncertainty, our SARL augments the asset information with their price movement prediction as additional states, where the prediction can be solely based on financial data (e.g., asset prices) or derived from alternative sources such as news. Experiments on two real-world datasets, (i) Bitcoin market and (ii) HighTech stock market with 7-year Reuters news articles, validate the effectiveness of SARL over existing PM approaches, both in terms of accumulated profits and risk-adjusted profits. Moreover, extensive simulations are conducted to demonstrate the importance of our proposed state augmentation, providing new insights and boosting performance significantly over standard RL-based PM method and other baselines.

q-fin.PM↗

arXiv2020

CIAN: Cross-Image Affinity Net for Weakly Supervised Semantic Segmentation

Weakly supervised semantic segmentation with only image-level labels saves large human effort to annotate pixel-level labels. Cutting-edge approaches rely on various innovative constraints and heuristic rules to generate the masks for every single image. Although great progress has been achieved by these methods, they treat each image independently and do not take account of the relationships across different images. In this paper, however, we argue that the cross-image relationship is vital for weakly supervised segmentation. Because it connects related regions across images, where supplementary representations can be propagated to obtain more consistent and integral regions. To leverage this information, we propose an end-to-end cross-image affinity module, which exploits pixel-level cross-image relationships with only image-level labels. By means of this, our approach achieves 64.3% and 65.3% mIoU on Pascal VOC 2012 validation and test set respectively, which is a new state-of-the-art result by only using image-level labels for weakly supervised semantic segmentation, demonstrating the superiority of our approach.

cs.CV↗

arXiv2019

A low-energy compact Shanghai-Wuhan electron beam ion trap for extraction of highly charged ions

A low-energy, compact and superconducting electron beam ion trap (the Shanghai-Wuhan EBIT or SW-EBIT) for extraction of highly charged ions is presented. The magnetic field in the central drift tube of the SW-EBIT is approximately 0.21 T produced by a pair of high-temperature superconducting coils. The electron-beam energy of the SW-EBIT is in the range of 30-4000 eV, and the maximum electron-beam current is up to 9 mA. Acting as a source of highly charged ions, the ion-beam optics for extraction is integrated, including an ion extractor and an einzel lens. A Wien filter is then used to measure the charge-state distribution of the extracted ions. In this work, the tungsten ions below the charge state of 15 have been produced, extracted, and analyzed. The charge-state distributions and spectra in the range of 530-580 nm of tungsten ions have been measured simultaneously with the electron-beam energy of 279 eV and 300 eV, which preliminarily indicates that the 549.9 nm line comes from $W^{14+}$.

physics.atom-ph↗

arXiv2019

Counterfactual Critic Multi-Agent Training for Scene Graph Generation

Scene graphs -- objects as nodes and visual relationships as edges -- describe the whereabouts and interactions of the things and stuff in an image for comprehensive scene understanding. To generate coherent scene graphs, almost all existing methods exploit the fruitful visual context by modeling message passing among objects, fitting the dynamic nature of reasoning with visual context, eg, "person" on "bike" can help to determine the relationship "ride", which in turn contributes to the category confidence of the two objects. However, we argue that the scene dynamics is not properly learned by using the prevailing cross-entropy based supervised learning paradigm, which is not sensitive to graph inconsistency: errors at the hub or non-hub nodes are unfortunately penalized equally. To this end, we propose a Counterfactual critic Multi-Agent Training (CMAT) approach to resolve the mismatch. CMAT is a multi-agent policy gradient method that frames objects as cooperative agents, and then directly maximizes a graph-level metric as the reward. In particular, to assign the reward properly to each agent, CMAT uses a counterfactual baseline that disentangles the agent-specific reward by fixing the dynamics of other agents. Extensive validations on the challenging Visual Genome benchmark show that CMAT achieves a state-of-the-art by significant performance gains under various settings and metrics.

cs.CV↗

arXiv2019

Weak Supervision Enhanced Generative Network for Question Generation

Automatic question generation according to an answer within the given passage is useful for many applications, such as question answering system, dialogue system, etc. Current neural-based methods mostly take two steps which extract several important sentences based on the candidate answer through manual rules or supervised neural networks and then use an encoder-decoder framework to generate questions about these sentences. These approaches neglect the semantic relations between the answer and the context of the whole passage which is sometimes necessary for answering the question. To address this problem, we propose the Weak Supervision Enhanced Generative Network (WeGen) which automatically discovers relevant features of the passage given the answer span in a weakly supervised manner to improve the quality of generated questions. More specifically, we devise a discriminator, Relation Guider, to capture the relations between the whole passage and the associated answer and then the Multi-Interaction mechanism is deployed to transfer the knowledge dynamically for our question generation system. Experiments show the effectiveness of our method in both automatic evaluations and human evaluations.

cs.CL↗

arXiv2019

Galaxy Learning -- A Position Paper

The recent rapid development of artificial intelligence (AI, mainly driven by machine learning research, especially deep learning) has achieved phenomenal success in various applications. However, to further apply AI technologies in real-world context, several significant issues regarding the AI ecosystem should be addressed. We identify the main issues as data privacy, ownership, and exchange, which are difficult to be solved with the current centralized paradigm of machine learning training methodology. As a result, we propose a novel model training paradigm based on blockchain, named Galaxy Learning, which aims to train a model with distributed data and to reserve the data ownership for their owners. In this new paradigm, encrypted models are moved around instead, and are federated once trained. Model training, as well as the communication, is achieved with blockchain and its smart contracts. Pricing of training data is determined by its contribution, and therefore it is not about the exchange of data ownership. In this position paper, we describe the motivation, paradigm, design, and challenges as well as opportunities of Galaxy Learning.

cs.OH↗

arXiv2018

Textually Guided Ranking Network for Attentional Image Retweet Modeling

Retweet prediction is a challenging problem in social media sites (SMS). In this paper, we study the problem of image retweet prediction in social media, which predicts the image sharing behavior that the user reposts the image tweets from their followees. Unlike previous studies, we learn user preference ranking model from their past retweeted image tweets in SMS. We first propose heterogeneous image retweet modeling network (IRM) that exploits users' past retweeted image tweets with associated contexts, their following relations in SMS and preference of their followees. We then develop a novel attentional multi-faceted ranking network learning framework with textually guided multi-modal neural networks for the proposed heterogenous IRM network to learn the joint image tweet representations and user preference representations for prediction task. The extensive experiments on a large-scale dataset from Twitter site shows that our method achieves better performance than other state-of-the-art solutions to the problem.

cs.IR↗

arXiv2018

Zero-Shot Visual Recognition using Semantics-Preserving Adversarial Embedding Networks

We propose a novel framework called Semantics-Preserving Adversarial Embedding Network (SP-AEN) for zero-shot visual recognition (ZSL), where test images and their classes are both unseen during training. SP-AEN aims to tackle the inherent problem --- semantic loss --- in the prevailing family of embedding-based ZSL, where some semantics would be discarded during training if they are non-discriminative for training classes, but could become critical for recognizing test classes. Specifically, SP-AEN prevents the semantic loss by introducing an independent visual-to-semantic space embedder which disentangles the semantic space into two subspaces for the two arguably conflicting objectives: classification and reconstruction. Through adversarial learning of the two subspaces, SP-AEN can transfer the semantics from the reconstructive subspace to the discriminative one, accomplishing the improved zero-shot recognition of unseen classes. Comparing with prior works, SP-AEN can not only improve classification but also generate photo-realistic images, demonstrating the effectiveness of semantic preservation. On four popular benchmarks: CUB, AWA, SUN and aPY, SP-AEN considerably outperforms other state-of-the-art methods by an absolute performance difference of 12.2\%, 9.3\%, 4.0\%, and 3.6\% in terms of harmonic mean values

cs.CV↗

arXiv2017

Attentional Factorization Machines: Learning the Weight of Feature Interactions via Attention Networks

Factorization Machines (FMs) are a supervised learning approach that enhances the linear regression model by incorporating the second-order feature interactions. Despite effectiveness, FM can be hindered by its modelling of all feature interactions with the same weight, as not all feature interactions are equally useful and predictive. For example, the interactions with useless features may even introduce noises and adversely degrade the performance. In this work, we improve FM by discriminating the importance of different feature interactions. We propose a novel model named Attentional Factorization Machine (AFM), which learns the importance of each feature interaction from data via a neural attention network. Extensive experiments on two real-world datasets demonstrate the effectiveness of AFM. Empirically, it is shown on regression task AFM betters FM with a $8.6\%$ relative improvement, and consistently outperforms the state-of-the-art deep learning methods Wide&Deep and DeepCross with a much simpler structure and fewer model parameters. Our implementation of AFM is publicly available at: https://github.com/hexiangnan/attentional_factorization_machine

cs.LG↗

arXiv2017

Graph-Theoretic Spatiotemporal Context Modeling for Video Saliency Detection

As an important and challenging problem in computer vision, video saliency detection is typically cast as a spatiotemporal context modeling problem over consecutive frames. As a result, a key issue in video saliency detection is how to effectively capture the intrinsical properties of atomic video structures as well as their associated contextual interactions along the spatial and temporal dimensions. Motivated by this observation, we propose a graph-theoretic video saliency detection approach based on adaptive video structure discovery, which is carried out within a spatiotemporal atomic graph. Through graph-based manifold propagation, the proposed approach is capable of effectively modeling the semantically contextual interactions among atomic video structures for saliency detection while preserving spatial smoothness and temporal consistency. Experiments demonstrate the effectiveness of the proposed approach over several benchmark datasets.

cs.CV↗

arXiv2017

Video Question Answering via Attribute-Augmented Attention Network Learning

Video Question Answering is a challenging problem in visual information retrieval, which provides the answer to the referenced video content according to the question. However, the existing visual question answering approaches mainly tackle the problem of static image question, which may be ineffectively for video question answering due to the insufficiency of modeling the temporal dynamics of video contents. In this paper, we study the problem of video question answering by modeling its temporal dynamics with frame-level attention mechanism. We propose the attribute-augmented attention network learning framework that enables the joint frame-level attribute detection and unified video representation learning for video question answering. We then incorporate the multi-step reasoning process for our proposed attention network to further improve the performance. We construct a large-scale video question answering dataset. We conduct the experiments on both multiple-choice and open-ended video question answering tasks to show the effectiveness of the proposed method.

cs.CV↗

arXiv2017

SCA-CNN: Spatial and Channel-wise Attention in Convolutional Networks for Image Captioning

Visual attention has been successfully applied in structural prediction tasks such as visual captioning and question answering. Existing visual attention models are generally spatial, i.e., the attention is modeled as spatial probabilities that re-weight the last conv-layer feature map of a CNN encoding an input image. However, we argue that such spatial attention does not necessarily conform to the attention mechanism --- a dynamic feature extractor that combines contextual fixations over time, as CNN features are naturally spatial, channel-wise and multi-layer. In this paper, we introduce a novel convolutional neural network dubbed SCA-CNN that incorporates Spatial and Channel-wise Attentions in a CNN. In the task of image captioning, SCA-CNN dynamically modulates the sentence generation context in multi-layer feature maps, encoding where (i.e., attentive spatial locations at multiple layers) and what (i.e., attentive channels) the visual attention is. We evaluate the proposed SCA-CNN architecture on three benchmark image captioning datasets: Flickr8K, Flickr30K, and MSCOCO. It is consistently observed that SCA-CNN significantly outperforms state-of-the-art visual attention-based image captioning methods.

cs.CV↗

arXiv2016

Ultrafast fluorescent decay induced by metal-mediated dipole-dipole interaction in two-dimensional molecular aggregates

Two-dimensional molecular aggregate (2DMA), a thin sheet of strongly interacting dipole molecules self-assembled at close distance on an ordered lattice, is a fascinating fluorescent material. It is distinctively different from the single or colloidal dye molecules or quantum dots in most previous research. In this paper, we verify for the first time that when a 2DMA is placed at a nanometric distance from a metallic substrate, the strong and coherent interaction between the dipoles inside the 2DMA dominates its fluorescent decay at picosecond timescale. Our streak-camera lifetime measurement and interacting lattice-dipole calculation reveal that the metal-mediated dipole-dipole interaction shortens the fluorescent lifetime to about one half and increases the energy dissipation rate by ten times than expected from the noninteracting single-dipole picture. Our finding can enrich our understanding of nanoscale energy transfer in molecular excitonic systems and may designate a new direction for developing fast and efficient optoelectronic devices.

physics.optics↗

arXiv2015

Tungsten spectroscopy in the EUV observed in SH-HtscEBIT

We have recorded extreme ultraviolet spectra from $\mathrm{W^{11+}}$ to $\mathrm{W^{15+}}$ ions using a new flat field spectrometer installed at the Shanghai high temperature superconducting electron beam ion trap. The spectra were recorded at beam energies ranging between 200 eV and 400 eV and showed spectral lines/transition arrays in the 170 - 260 Å region. The charge states and spectra transitions were identified by comparison with calculations using a detailed relativistic configuration interaction method and collisional-radiative model, both incorporated in the Flexible Atomic Code. Atomic structure calculations showed that the dominant emission arises from $5d$ $\rightarrow$ $5p$ and $5p$ $\rightarrow$ $5s$ transitions. The work also identified the ground-state configuration of $W^{13+}$ as $4f^{13}5s^2$ both theoretically and experimentally.

physics.atom-ph↗

arXiv2015

Optical Selection Rule based on Valley-Exciton Locking for 2D Valleytronics

Optical selection rule fundamentally determines the optical transition between energy states in a variety of physical systems from hydrogen atoms to bulk crystals such as GaAs. It is important for optoelectronic applications such as lasers, energy-dispersive X-ray spectroscopy and quantum computation. Recently, single layer transition metal dichalcogenide (TMDC) exhibits valleys in momentum space with nontrivial Berry curvature and excitons with large binding energy. However, it is unclear how the unique valley degree of freedom combined with the strong excitonic effect influences the optical excitation. Here we discover a new set of optical selection rules in monolayer WS2,imposed by valley and exciton angular momentum. We experimentally demonstrated such a principle for second harmonic generation (SHG) and two-photon luminescence (TPL). Moreover, the two-photon induced valley populations yield net circular polarized photoluminescence after a sub-ps interexciton relaxation (2p->1s) and last for 8 ps. The discovery of this new optical selection rule in valleytronic 2D system not only largely extend information degrees but sets a foundation in control of optical transitions that is crucial to valley optoeletronic device applications such as 2D valley-polarized light emitting diodes (LED), optical switches and coherent control for quantum computing.

cond-mat.mtrl-sci↗

arXiv2014

Metric Learning Driven Multi-Task Structured Output Optimization for Robust Keypoint Tracking

As an important and challenging problem in computer vision and graphics, keypoint-based object tracking is typically formulated in a spatio-temporal statistical learning framework. However, most existing keypoint trackers are incapable of effectively modeling and balancing the following three aspects in a simultaneous manner: temporal model coherence across frames, spatial model consistency within frames, and discriminative feature construction. To address this issue, we propose a robust keypoint tracker based on spatio-temporal multi-task structured output optimization driven by discriminative metric learning. Consequently, temporal model coherence is characterized by multi-task structured keypoint model learning over several adjacent frames, while spatial model consistency is modeled by solving a geometric verification based structured learning problem. Discriminative feature construction is enabled by metric learning to ensure the intra-class compactness and inter-class separability. Finally, the above three modules are simultaneously optimized in a joint learning scheme. Experimental results have demonstrated the effectiveness of our tracker.

cs.CV↗