arXiv subjects

Zhiqiang Shen

Publications and source records attributed to Zhiqiang Shen.

At least 199 records · Page 11Linked to original sources

arXiv2020

Conditional Link Prediction of Category-Implicit Keypoint Detection

Keypoints of objects reflect their concise abstractions, while the corresponding connection links (CL) build the skeleton by detecting the intrinsic relations between keypoints. Existing approaches are typically computationally-intensive, inapplicable for instances belonging to multiple classes, and/or infeasible to simultaneously encode connection information. To address the aforementioned issues, we propose an end-to-end category-implicit Keypoint and Link Prediction Network (KLPNet), which is the first approach for simultaneous semantic keypoint detection (for multi-class instances) and CL rejuvenation. In our KLPNet, a novel Conditional Link Prediction Graph is proposed for link prediction among keypoints that are contingent on a predefined category. Furthermore, a Cross-stage Keypoint Localization Module (CKLM) is introduced to explore feature aggregation for coarse-to-fine keypoint localization. Comprehensive experiments conducted on three publicly available benchmarks demonstrate that our KLPNet consistently outperforms all other state-of-the-art approaches. Furthermore, the experimental results of CL prediction also show the effectiveness of our KLPNet with respect to occlusion problems.

cs.CV↗

arXiv2020

Online Ensemble Model Compression using Knowledge Distillation

This paper presents a novel knowledge distillation based model compression framework consisting of a student ensemble. It enables distillation of simultaneously learnt ensemble knowledge onto each of the compressed student models. Each model learns unique representations from the data distribution due to its distinct architecture. This helps the ensemble generalize better by combining every model's knowledge. The distilled students and ensemble teacher are trained simultaneously without requiring any pretrained weights. Moreover, our proposed method can deliver multi-compressed students with single training, which is efficient and flexible for different scenarios. We provide comprehensive experiments using state-of-the-art classification models to validate our framework's effectiveness. Notably, using our framework a 97% compressed ResNet110 student model managed to produce a 10.64% relative accuracy gain over its individual baseline training on CIFAR100 dataset. Similarly a 95% compressed DenseNet-BC(k=12) model managed a 8.17% relative accuracy gain.

cs.CV↗

arXiv2020

Channel-wise Alignment for Adaptive Object Detection

Generic object detection has been immensely promoted by the development of deep convolutional neural networks in the past decade. However, in the domain shift circumstance, the changes in weather, illumination, etc., often cause domain gap, and thus performance drops substantially when detecting objects from one domain to another. Existing methods on this task usually draw attention on the high-level alignment based on the whole image or object of interest, which naturally, cannot fully utilize the fine-grained channel information. In this paper, we realize adaptation from a thoroughly different perspective, i.e., channel-wise alignment. Motivated by the finding that each channel focuses on a specific pattern (e.g., on special semantic regions, such as car), we aim to align the distribution of source and target domain on the channel level, which is finer for integration between discrepant domains. Our method mainly consists of self channel-wise and cross channel-wise alignment. These two parts explore the inner-relation and cross-relation of attention regions implicitly from the view of channels. Further more, we also propose a RPN domain classifier module to obtain a domain-invariant RPN network. Extensive experiments show that the proposed method performs notably better than existing methods with about 5% improvement under various domain-shift settings. Experiments on different task (e.g. instance segmentation) also demonstrate its good scalability.

cs.CV↗

arXiv2020

Solving Missing-Annotation Object Detection with Background Recalibration Loss

This paper focuses on a novel and challenging detection scenario: A majority of true objects/instances is unlabeled in the datasets, so these missing-labeled areas will be regarded as the background during training. Previous art on this problem has proposed to use soft sampling to re-weight the gradients of RoIs based on the overlaps with positive instances, while their method is mainly based on the two-stage detector (i.e. Faster RCNN) which is more robust and friendly for the missing label scenario. In this paper, we introduce a superior solution called Background Recalibration Loss (BRL) that can automatically re-calibrate the loss signals according to the pre-defined IoU threshold and input image. Our design is built on the one-stage detector which is faster and lighter. Inspired by the Focal Loss formulation, we make several significant modifications to fit on the missing-annotation circumstance. We conduct extensive experiments on the curated PASCAL VOC and MS COCO datasets. The results demonstrate that our proposed method outperforms the baseline and other state-of-the-arts by a large margin. Code available: https://github.com/Dwrety/mmdetection-selective-iou.

cs.CV↗

arXiv2020

ReActNet: Towards Precise Binary Neural Network with Generalized Activation Functions

In this paper, we propose several ideas for enhancing a binary network to close its accuracy gap from real-valued networks without incurring any additional computational cost. We first construct a baseline network by modifying and binarizing a compact real-valued network with parameter-free shortcuts, bypassing all the intermediate convolutional layers including the downsampling layers. This baseline network strikes a good trade-off between accuracy and efficiency, achieving superior performance than most of existing binary networks at approximately half of the computational cost. Through extensive experiments and analysis, we observed that the performance of binary networks is sensitive to activation distribution variations. Based on this important observation, we propose to generalize the traditional Sign and PReLU functions, denoted as RSign and RPReLU for the respective generalized functions, to enable explicit learning of the distribution reshape and shift at near-zero extra cost. Lastly, we adopt a distributional loss to further enforce the binary network to learn similar output distributions as those of a real-valued network. We show that after incorporating all these ideas, the proposed ReActNet outperforms all the state-of-the-arts by a large margin. Specifically, it outperforms Real-to-Binary Net and MeliusNet29 by 4.0% and 3.6% respectively for the top-1 accuracy and also reduces the gap to its real-valued counterpart to within 3.0% top-1 accuracy on ImageNet dataset. Code and models are available at: https://github.com/liuzechun/ReActNet.

cs.CV↗

arXiv2020

Soft Anchor-Point Object Detection

Recently, anchor-free detection methods have been through great progress. The major two families, anchor-point detection and key-point detection, are at opposite edges of the speed-accuracy trade-off, with anchor-point detectors having the speed advantage. In this work, we boost the performance of the anchor-point detector over the key-point counterparts while maintaining the speed advantage. To achieve this, we formulate the detection problem from the anchor point's perspective and identify ineffective training as the main problem. Our key insight is that anchor points should be optimized jointly as a group both within and across feature pyramid levels. We propose a simple yet effective training strategy with soft-weighted anchor points and soft-selected pyramid levels to address the false attention issue within each pyramid level and the feature selection issue across all the pyramid levels, respectively. To evaluate the effectiveness, we train a single-stage anchor-free detector called Soft Anchor-Point Detector (SAPD). Experiments show that our concise SAPD pushes the envelope of speed/accuracy trade-off to a new level, outperforming recent state-of-the-art anchor-free and anchor-based detectors. Without bells and whistles, our best model can achieve a single-model single-scale AP of 47.4% on COCO.

cs.CV↗

arXiv2020

Cross-Supervised Object Detection

After learning a new object category from image-level annotations (with no object bounding boxes), humans are remarkably good at precisely localizing those objects. However, building good object localizers (i.e., detectors) currently requires expensive instance-level annotations. While some work has been done on learning detectors from weakly labeled samples (with only class labels), these detectors do poorly at localization. In this work, we show how to build better object detectors from weakly labeled images of new categories by leveraging knowledge learned from fully labeled base categories. We call this novel learning paradigm cross-supervised object detection. We propose a unified framework that combines a detection head trained from instance-level annotations and a recognition head learned from image-level annotations, together with a spatial correlation module that bridges the gap between detection and recognition. These contributions enable us to better detect novel objects with image-level annotations in complex multi-object scenes such as the COCO dataset.

cs.CV↗

arXiv2020

Space VLBI 2020: Science and Technology Futures Conference Summary

The "Space VLBI 2020: Science and Technology Futures" meeting was the second in The Future of High-Resolution Radio Interferometry in Space series. The first meeting (2018 September 5--6; Noordwijk, the Netherlands) focused on the full range of science applications possible for very long baseline interferometry (VLBI) with space-based antennas. Accordingly, the observing frequencies (wavelengths) considered ranged from below 1~MHz (> 300 m) to above 300~GHz (< 1 mm). For this second meeting, the focus was narrowed to mission concepts and the supporting technologies to enable the highest angular resolution observations at frequencies of 30~GHz and higher (< 1 cm). This narrowing of focus was driven by both scientific and technical considerations. First, results from the RadioAstron mission and the Event Horizon Telescope (EHT) have generated considerable excitement for studying the inner portions of black hole (BH) accretion disks and jets and testing elements of the General Theory of Relativity (GR). Second, the technologies and requirements involved in space-based VLBI differ considerably between 100~MHz and 100~GHz; a related consideration is that there are a number of existing instruments or mission concepts for frequencies of approximately 100~MHz and below, while it has been some time since attention has been devoted to space VLBI at frequencies above 10~GHz. This conference summary attempts to capture elements of presentations and discussions that occurred.

astro-ph.IM↗

arXiv2020

Binarizing MobileNet via Evolution-based Searching

Binary Neural Networks (BNNs), known to be one among the effectively compact network architectures, have achieved great outcomes in the visual tasks. Designing efficient binary architectures is not trivial due to the binary nature of the network. In this paper, we propose a use of evolutionary search to facilitate the construction and training scheme when binarizing MobileNet, a compact network with separable depth-wise convolution. Inspired by one-shot architecture search frameworks, we manipulate the idea of group convolution to design efficient 1-Bit Convolutional Neural Networks (CNNs), assuming an approximately optimal trade-off between computational cost and model accuracy. Our objective is to come up with a tiny yet efficient binary neural architecture by exploring the best candidates of the group convolution while optimizing the model performance in terms of complexity and latency. The approach is threefold. First, we train strong baseline binary networks with a wide range of random group combinations at each convolutional layer. This set-up gives the binary neural networks a capability of preserving essential information through layers. Second, to find a good set of hyperparameters for group convolutions we make use of the evolutionary search which leverages the exploration of efficient 1-bit models. Lastly, these binary models are trained from scratch in a usual manner to achieve the final binary model. Various experiments on ImageNet are conducted to show that following our construction guideline, the final model achieves 60.09% Top-1 accuracy and outperforms the state-of-the-art CI-BCNN with the same computational cost.

cs.CV↗

arXiv2020

The radio properties of the OH megamaser galaxy IRAS 02524+2046

We present results from VLBI observations of continuum and OH line emission in IRAS 02524+2046 and also arcsecond-scale radio properties of this galaxy using VLA archive data. We found that there is no significant detection of radio continuum emission from VLBI observations. The arcsecond-scale radio images of this source show no clear extended emission, the total radio flux density at L and C band are around 2.9 mJy and 1.0 mJy respectively, which indicate a steep radio spectral index between the two band. Steep spectral index, low brightness temperature and high $q$-ratio (the FIR to the radio flux density), which are three critical indicators in classification of radio activity in the nuclei of galaxies, are all consistent with the classification of this source as a starburst galaxy from its optical spectrum. The high-resolution line profile show that both of \textbf{the 1665 and 1667 MHz OH maser} line have been detected which show three and two clear components respectively. The channel maps show that the maser emission are distributed in a region $\sim$ 210 pc $\times$ 90 pc, the detected maser components at different region show similar double spectral feature, which might be an evidence that this galaxy is at a stage of major merger as seen from the optical morphology.

astro-ph.GA↗

arXiv2020

Attentive CutMix: An Enhanced Data Augmentation Approach for Deep Learning Based Image Classification

Convolutional neural networks (CNN) are capable of learning robust representation with different regularization methods and activations as convolutional layers are spatially correlated. Based on this property, a large variety of regional dropout strategies have been proposed, such as Cutout, DropBlock, CutMix, etc. These methods aim to promote the network to generalize better by partially occluding the discriminative parts of objects. However, all of them perform this operation randomly, without capturing the most important region(s) within an object. In this paper, we propose Attentive CutMix, a naturally enhanced augmentation strategy based on CutMix. In each training iteration, we choose the most descriptive regions based on the intermediate attention maps from a feature extractor, which enables searching for the most discriminative parts in an image. Our proposed method is simple yet effective, easy to implement and can boost the baseline significantly. Extensive experiments on CIFAR-10/100, ImageNet datasets with various CNN architectures (in a unified setting) demonstrate the effectiveness of our proposed method, which consistently outperforms the baseline CutMix and other methods by a significant margin.

cs.CV↗

arXiv2019

Mapping Observations of complex organic molecules around Sagittarius B2 with ARO 12m telescope

We performed high-sensitivity mapping observations of several complex organic molecules around Sagittarius B2 with ARO 12m telescope at 3-mm wavelength. Based on their spatial distribution, molecules can be classified as either "extended" that detected not only in Sgr B2(N) and Sgr B2(M), or "compact" that only detected toward or near to Sgr B2(N) and Sgr B2(M). The "extended" molecules including glycolaldehyde (CH2OHCHO), methyl formate (CH3OCHO), formic acid (t-HCOOH), ethanol (C2H5OH) and methyl amine (CH3NH2), while the "compact" molecules including dimethyl ether (CH3OCH3), ethyl cyanide (C2H5CN), and amino acetonitrile (H2NCH2CN). These "compact" molecules are likely produced under strong UV radiation, while "extended" molecules are likely formed under low-temperature, via gas-phase or grain surface reactions. The spatial distribution of "warm" CH2OHCHO at 89 GHz differ from the spatial distribution of "cold" CH2OHCHO observed at 13 GHz. We found evidence for an overabundance of CH2OHCHO compared to that expected from the gas-phase model, which indicates that grain-surface reactions are necessary to explain the origin of CH2OHCHO in Sagittarius B2. Grain-surface reactions are also needed to explain the correlation between the abundances of "cold" CH2OHCHO and C2H5OH. These results demonstrate the importance of grain-surface chemistry in the production of complex organic molecules.

astro-ph.SR↗

arXiv2019

SCL: Towards Accurate Domain Adaptive Object Detection via Gradient Detach Based Stacked Complementary Losses

Unsupervised domain adaptive object detection aims to learn a robust detector in the domain shift circumstance, where the training (source) domain is label-rich with bounding box annotations, while the testing (target) domain is label-agnostic and the feature distributions between training and testing domains are dissimilar or even totally different. In this paper, we propose a gradient detach based stacked complementary losses (SCL) method that uses detection losses as the primary objective, and cuts in several auxiliary losses in different network stages accompanying with gradient detach training to learn more discriminative representations. We argue that the prior methods mainly leverage more loss functions for training but ignore the interaction of different losses and also the compatible training strategy (gradient detach updating in our work). Thus, our proposed method is a more syncretic adaptation learning process. We conduct comprehensive experiments on seven datasets, the results demonstrate that our method performs favorably better than the state-of-the-art methods by a significant margin. For instance, from Cityscapes to FoggyCityscapes, we achieve 37.9% mAP, outperforming the previous art Strong-Weak by 3.6%.

cs.CV↗

arXiv2019

Kinematics of the M87 jet in the collimation zone: gradual acceleration and velocity stratification

We study the kinematics of the M87 jet using the first year data of the KVN and VERA Array (KaVA) large program, which has densely monitored the jet at 22 and 43 GHz since 2016. We find that the apparent jet speeds generally increase from $\approx0.3c$ at $\approx0.5$ mas from the jet base to $\approx2.7c$ at $\approx20$ mas, indicating that the jet is accelerated from subluminal to superluminal speeds on these scales. We perform a complementary jet kinematic analysis by using archival Very Long Baseline Array monitoring data observed in $2005-2009$ at 1.7 GHz and find that the jet is moving at relativistic speeds up to $\approx5.8c$ at distances of $200-410$ mas. We combine the two kinematic results and find that the jet is gradually accelerated over a broad distance range that coincides with the jet collimation zone, implying that conversion of Poynting flux to kinetic energy flux takes place. If the jet emission consists of a single streamline, the observed trend of jet acceleration ($Γ\propto z^{0.16\pm0.01}$) is relatively slow compared to models of a highly magnetized jet. This indicates that Poynting flux conversion through the differential collimation of poloidal magnetic fields may be less efficient than expected. However, we find a non-negligible dispersion in the observed speeds for a given jet distance, making it difficult to describe the jet velocity field with a single power-law acceleration function. We discuss the possibility that the jet emission consists of multiple streamlines following different acceleration profiles, resulting in jet velocity stratification.

astro-ph.HE↗

arXiv2019

Adversarial-Based Knowledge Distillation for Multi-Model Ensemble and Noisy Data Refinement

Generic Image recognition is a fundamental and fairly important visual problem in computer vision. One of the major challenges of this task lies in the fact that single image usually has multiple objects inside while the labels are still one-hot, another one is noisy and sometimes missing labels when annotated by humans. In this paper, we focus on tackling these challenges accompanying with two different image recognition problems: multi-model ensemble and noisy data recognition with a unified framework. As is well-known, usually the best performing deep neural models are ensembles of multiple base-level networks, as it can mitigate the variation or noise containing in the dataset. Unfortunately, the space required to store these many networks, and the time required to execute them at runtime, prohibit their use in applications where test sets are large (e.g., ImageNet). In this paper, we present a method for compressing large, complex trained ensembles into a single network, where the knowledge from a variety of trained deep neural networks (DNNs) is distilled and transferred to a single DNN. In order to distill diverse knowledge from different trained (teacher) models, we propose to use adversarial-based learning strategy where we define a block-wise training loss to guide and optimize the predefined student network to recover the knowledge in teacher models, and to promote the discriminator network to distinguish teacher vs. student features simultaneously. Extensive experiments on CIFAR-10/100, SVHN, ImageNet and iMaterialist Challenge Dataset demonstrate the effectiveness of our MEAL method. On ImageNet, our ResNet-50 based MEAL achieves top-1/5 21.79%/5.99% val error, which outperforms the original model by 2.06%/1.14%. On iMaterialist Challenge Dataset, our MEAL obtains a remarkable improvement of top-3 1.15% (official evaluation metric) on a strong baseline model of ResNet-101.

cs.CV↗

arXiv2019

MoBiNet: A Mobile Binary Network for Image Classification

MobileNet and Binary Neural Networks are two among the most widely used techniques to construct deep learning models for performing a variety of tasks on mobile and embedded platforms.In this paper, we present a simple yet efficient scheme to exploit MobileNet binarization at activation function and model weights. However, training a binary network from scratch with separable depth-wise and point-wise convolutions in case of MobileNet is not trivial and prone to divergence. To tackle this training issue, we propose a novel neural network architecture, namely MoBiNet - Mobile Binary Network in which skip connections are manipulated to prevent information loss and vanishing gradient, thus facilitate the training process. More importantly, while existing binary neural networks often make use of cumbersome backbones such as Alex-Net, ResNet, VGG-16 with float-type pre-trained weights initialization, our MoBiNet focuses on binarizing the already-compressed neural networks like MobileNet without the need of a pre-trained model to start with. Therefore, our proposal results in an effectively small model while keeping the accuracy comparable to existing ones. Experiments on ImageNet dataset show the potential of the MoBiNet as it achieves 54.40% top-1 accuracy and dramatically reduces the computational cost with binary operators.

cs.CV↗

arXiv2019

MEAL: Multi-Model Ensemble via Adversarial Learning

Often the best performing deep neural models are ensembles of multiple base-level networks. Unfortunately, the space required to store these many networks, and the time required to execute them at test-time, prohibits their use in applications where test sets are large (e.g., ImageNet). In this paper, we present a method for compressing large, complex trained ensembles into a single network, where knowledge from a variety of trained deep neural networks (DNNs) is distilled and transferred to a single DNN. In order to distill diverse knowledge from different trained (teacher) models, we propose to use adversarial-based learning strategy where we define a block-wise training loss to guide and optimize the predefined student network to recover the knowledge in teacher models, and to promote the discriminator network to distinguish teacher vs. student features simultaneously. The proposed ensemble method (MEAL) of transferring distilled knowledge with adversarial learning exhibits three important advantages: (1) the student network that learns the distilled knowledge with discriminators is optimized better than the original model; (2) fast inference is realized by a single forward pass, while the performance is even better than traditional ensembles from multi-original models; (3) the student network can learn the distilled knowledge from a teacher model that has arbitrary structures. Extensive experiments on CIFAR-10/100, SVHN and ImageNet datasets demonstrate the effectiveness of our MEAL method. On ImageNet, our ResNet-50 based MEAL achieves top-1/5 21.79%/5.99% val error, which outperforms the original model by 2.06%/1.14%. Code and models are available at: https://github.com/AaronHeee/MEAL

cs.CV↗

arXiv2019

Improving Object Detection from Scratch via Gated Feature Reuse

In this paper, we present a simple and parameter-efficient drop-in module for one-stage object detectors like SSD when learning from scratch (i.e., without pre-trained models). We call our module GFR (Gated Feature Reuse), which exhibits two main advantages. First, we introduce a novel gate-controlled prediction strategy enabled by Squeeze-and-Excitation to adaptively enhance or attenuate supervision at different scales based on the input object size. As a result, our model is more effective in detecting diverse sizes of objects. Second, we propose a feature-pyramids structure to squeeze rich spatial and semantic features into a single prediction layer, which strengthens feature representation and reduces the number of parameters to learn. We apply the proposed structure on DSOD and SSD detection frameworks, and evaluate the performance on PASCAL VOC 2007, 2012 and COCO datasets. With fewer model parameters, GFR-DSOD outperforms the baseline DSOD by 1.4%, 1.1%, 1.7% and 0.6%, respectively. GFR-SSD also outperforms the original SSD and SSD with dense prediction by 3.6% and 2.8% on VOC 2007 dataset. Code is available at: https://github.com/szq0214/GFR-DSOD .

cs.CV↗