arXiv subjects

Chen Sun

Publications and source records attributed to Chen Sun.

At least 217 records · Page 12Linked to original sources

arXiv2019

Unsupervised Discovery of Parts, Structure, and Dynamics

Humans easily recognize object parts and their hierarchical structure by watching how they move; they can then predict how each part moves in the future. In this paper, we propose a novel formulation that simultaneously learns a hierarchical, disentangled object representation and a dynamics model for object parts from unlabeled videos. Our Parts, Structure, and Dynamics (PSD) model learns to, first, recognize the object parts via a layered image representation; second, predict hierarchy via a structural descriptor that composes low-level concepts into a hierarchical structure; and third, model the system dynamics by predicting the future. Experiments on multiple real and synthetic datasets demonstrate that our PSD model works well on all three tasks: segmenting object parts, building their hierarchical structure, and capturing their motion distributions.

cs.CV↗

arXiv2019

Stochastic Prediction of Multi-Agent Interactions from Partial Observations

We present a method that learns to integrate temporal information, from a learned dynamics model, with ambiguous visual information, from a learned vision model, in the context of interacting agents. Our method is based on a graph-structured variational recurrent neural network (Graph-VRNN), which is trained end-to-end to infer the current state of the (partially observed) world, as well as to forecast future states. We show that our method outperforms various baselines on two sports datasets, one based on real basketball trajectories, and one generated by a soccer game engine.

cs.LG↗

arXiv2019

Machine Learning Allows Calibration Models to Predict Trace Element Concentration in Soil with Generalized LIBS Spectra

Calibration models have been developed for determination of trace elements, silver for instance, in soil using laser-induced breakdown spectroscopy (LIBS). The major concern is the matrix effect. Although it affects the accuracy of LIBS measurements in a general way, the effect appears accentuated for soil because of large variation of chemical and physical properties among different soils. The purpose is to reduce its influence in such way an accurate and soil-independent calibration model can be constructed. At the same time, the developed model should efficiently reduce experimental fluctuations affecting measurement precision. A univariate model first reveals obvious influence of matrix effect and important experimental fluctuation. A multivariate model has been then developed. A key point is the introduction of generalized spectrum where variables representing the soil type are explicitly included. Machine learning has been used to develop the model. After a necessary pretreatment where a feature selection process reduces the dimension of raw spectrum accordingly to the number of available spectra, the data have been fed in to a back-propagation neuronal networks (BPNN) to train and validate the model. The resulted soilindependent calibration model allows average relative error of calibration (REC) and average relative error of prediction (REP) within the range of 5-6%.

physics.chem-ph↗

arXiv2019

D3D: Distilled 3D Networks for Video Action Recognition

State-of-the-art methods for video action recognition commonly use an ensemble of two networks: the spatial stream, which takes RGB frames as input, and the temporal stream, which takes optical flow as input. In recent work, both of these streams consist of 3D Convolutional Neural Networks, which apply spatiotemporal filters to the video clip before performing classification. Conceptually, the temporal filters should allow the spatial stream to learn motion representations, making the temporal stream redundant. However, we still see significant benefits in action recognition performance by including an entirely separate temporal stream, indicating that the spatial stream is "missing" some of the signal captured by the temporal stream. In this work, we first investigate whether motion representations are indeed missing in the spatial stream of 3D CNNs. Second, we demonstrate that these motion representations can be improved by distillation, by tuning the spatial stream to predict the outputs of the temporal stream, effectively combining both models into a single stream. Finally, we show that our Distilled 3D Network (D3D) achieves performance on par with two-stream approaches, using only a single model and with no need to compute optical flow.

cs.CV↗

arXiv2018

Composing Text and Image for Image Retrieval - An Empirical Odyssey

In this paper, we study the task of image retrieval, where the input query is specified in the form of an image plus some text that describes desired modifications to the input image. For example, we may present an image of the Eiffel tower, and ask the system to find images which are visually similar but are modified in small ways, such as being taken at nighttime instead of during the day. To tackle this task, we learn a similarity metric between a target image and a source image plus source text, an embedding and composing function such that target image feature is close to the source image plus text composition feature. We propose a new way to combine image and text using such function that is designed for the retrieval task. We show this outperforms existing approaches on 3 different datasets, namely Fashion-200k, MIT-States and a new synthetic dataset we create based on CLEVR. We also show that our approach can be used to classify input queries, in addition to image retrieval.

cs.CV↗

arXiv2018

Beam Domain Massive MIMO for Optical Wireless Communications with Transmit Lens

This paper presents a novel massive multiple-input multiple-output (MIMO) transmission in beam domain for optical wireless communications. The optical base station equipped with massive optical transmitters communicates with a number of user terminals (UTs) through a transmit lens. Focusing on LED transmitters, we analyze light refraction of the lens and establish a channel model for optical massive MIMO transmissions. For a large number of LEDs, channel vectors of different UTs become asymptotically orthogonal. We investigate the maximum ratio transmission and regularized zero-forcing precoding in the optical massive MIMO system, and propose a linear precoding design to maximize the sum rate. We further design the precoding when the number of transmitters grows asymptotically large, and show that beam division multiple access (BDMA) transmission achieves the asymptotically optimal performance for sum rate maximization. Unlike optical MIMO without a transmit lens, BDMA can increase the sum rate proportionally to $2K$ and $K$ under the total and per transmitter power constraints, respectively, where $K$ is the number of UTs. In the non-asymptotic case, we prove the orthogonality conditions of the optimal power allocation in beam domain and propose efficient beam allocation algorithms. Numerical results confirm the significantly improved performance of our proposed beam domain optical massive MIMO communication approaches.

eess.SP↗

arXiv2018

$B$-decay anomalies and scalar leptoquarks in unified Pati-Salam models from noncommutative geometry

Motivated by possible scalar-leptoquark explanations of the recently reported $B$-decay anomalies, we investigate whether the required leptoquarks can be accommodated within models based on noncommutative geometry (NCG). The models considered have the gauge structure of Pati-Salam models, $SU(4)\times SU(2)_L\times SU(2)_R$, with gauge coupling unification at a single scale. In one of the models, we find a unique scalar leptoquark with quantum numbers $(3,1,-\frac{1}{3})_{321}$, originating from a complex multiplet $(6,1,1)_{422}$, which can potentially explain the $B$-decay anomalies if its mass is on the order of a few TeV. The unification of couplings can be realized with the inclusion of a single step of intermediate symmetry breaking. The scalar leptoquark under consideration does not contribute to proton decay due to the absence of diquark couplings, as dictated by the underlying noncommutative geometry.

hep-ph↗

arXiv2018

CoCo: Compact and Optimized Consolidation of Modularized Service Function Chains in NFV

The modularization of Service Function Chains (SFCs) in Network Function Virtualization (NFV) could introduce significant performance overhead and resource efficiency degradation due to introducing frequent packet transfer and consuming much more hardware resources. In response, we exploit the lightweight and individually scalable features of elements in Modularized SFCs (MSFCs) and propose CoCo, a compact and optimized consolidation framework for MSFC in NFV. CoCo addresses the above problems in two ways. First, CoCo Optimized Placer pays attention to the problem of which elements to consolidate and provides a performance-aware placement algorithm to place MSFCs compactly and optimize the global packet transfer cost. Second, CoCo Individual Scaler innovatively introduces a push-aside scaling up strategy to avoid degrading performance and taking up new CPU cores. To support MSFC consolidation, CoCo also provides an automatic runtime scheduler to ensure fairness when elements are consolidated on CPU core. Our evaluation results show that CoCo achieves significant performance improvement and efficient resource utilization.

cs.NI↗

arXiv2018

Actor-Centric Relation Network

Current state-of-the-art approaches for spatio-temporal action localization rely on detections at the frame level and model temporal context with 3D ConvNets. Here, we go one step further and model spatio-temporal relations to capture the interactions between human actors, relevant objects and scene elements essential to differentiate similar human actions. Our approach is weakly supervised and mines the relevant elements automatically with an actor-centric relational network (ACRN). ACRN computes and accumulates pair-wise relation information from actor and global scene features, and generates relation features for action classification. It is implemented as neural networks and can be trained jointly with an existing action detection system. We show that ACRN outperforms alternative approaches which capture relation information, and that the proposed framework improves upon the state-of-the-art performance on JHMDB and AVA. A visualization of the learned relation features confirms that our approach is able to attend to the relevant relations for each action.

cs.CV↗

arXiv2018

Rethinking Spatiotemporal Feature Learning: Speed-Accuracy Trade-offs in Video Classification

Despite the steady progress in video analysis led by the adoption of convolutional neural networks (CNNs), the relative improvement has been less drastic as that in 2D static image classification. Three main challenges exist including spatial (image) feature representation, temporal information representation, and model/computation complexity. It was recently shown by Carreira and Zisserman that 3D CNNs, inflated from 2D networks and pretrained on ImageNet, could be a promising way for spatial and temporal representation learning. However, as for model/computation complexity, 3D CNNs are much more expensive than 2D CNNs and prone to overfit. We seek a balance between speed and accuracy by building an effective and efficient video classification system through systematic exploration of critical network design choices. In particular, we show that it is possible to replace many of the 3D convolutions by low-cost 2D convolutions. Rather surprisingly, best result (in both speed and accuracy) is achieved when replacing the 3D convolutions at the bottom of the network, suggesting that temporal representation learning on high-level semantic features is more useful. Our conclusion generalizes to datasets with very different properties. When combined with several other cost-effective designs including separable spatial/temporal convolution and feature gating, our system results in an effective video classification system that that produces very competitive results on several action classification benchmarks (Kinetics, Something-something, UCF101 and HMDB), as well as two action detection (localization) benchmarks (JHMDB and UCF101-24).

cs.CV↗

arXiv2018

PAM: When Overloaded, Push Your Neighbor Aside!

Recently SmartNICs are widely used to accelerate service chains in NFV. However, when the SmartNIC is overloaded, casually migrating vNFs away from SmartNIC to CPU may lead to additional packet transmissions between SmartNIC and CPU. To address this problem, we present PAM, push aside migration to effectively alleviate the hot spot on SmartNIC with no performance overhead. Our key novelty is to push vNFs on the border of SmartNIC and CPU aside to release resources for the bottleneck vNF. Evaluation shows that PAM could efficiently alleviate the hot spot on SmartNIC and generate a service chain with much lower latency compared with the naive solution.

cs.NI↗

arXiv2018

Large Scale Fine-Grained Categorization and Domain-Specific Transfer Learning

Transferring the knowledge learned from large scale datasets (e.g., ImageNet) via fine-tuning offers an effective solution for domain-specific fine-grained visual categorization (FGVC) tasks (e.g., recognizing bird species or car make and model). In such scenarios, data annotation often calls for specialized domain knowledge and thus is difficult to scale. In this work, we first tackle a problem in large scale FGVC. Our method won first place in iNaturalist 2017 large scale species classification challenge. Central to the success of our approach is a training scheme that uses higher image resolution and deals with the long-tailed distribution of training data. Next, we study transfer learning via fine-tuning from large scale datasets to small scale, domain-specific FGVC datasets. We propose a measure to estimate domain similarity via Earth Mover's Distance and demonstrate that transfer learning benefits from pre-training on a source domain that is similar to the target domain by this measure. Our proposed transfer learning outperforms ImageNet pre-training and obtains state-of-the-art results on multiple commonly used FGVC datasets.

cs.CV↗

arXiv2018

A large class of solvable multistate Landau-Zener models and quantum integrability

The concept of quantum integrability has been introduced recently for quantum systems with explicitly time-dependent Hamiltonians. Within the multistate Landau-Zener (MLZ) theory, however, there has been a successful alternative approach to identify and solve complex time-dependent models. Here we compare both methods by applying them to a new class of exactly solvable MLZ models. This class contains systems with an arbitrary number $N\ge 4$ of interacting states and shows a quickly growing with $N$ number of exact adiabatic energy crossing points, which appear at different moments of time. At each $N$, transition probabilities in these systems can be found analytically and exactly but complexity and variety of solutions in this class also grow with $N$ quickly. We illustrate how common features of solvable MLZ systems appear from quantum integrability and develop an approach to further classification of solvable MLZ problems.

quant-ph↗

arXiv2018

Hidden-Sector Spectroscopy with Gravitational Waves from Binary Neutron Stars

We show that neutron star binaries can be ideal laboratories to probe hidden sectors with a long range force. In particular, it is possible for gravitational wave detectors such as LIGO and Virgo to resolve the correction of waveforms from ultralight dark gauge bosons coupled to neutron stars. We observe that the interaction of the hidden sector affects both the gravitational wave frequency and amplitude in a way that cannot be fitted by pure gravity.

hep-ph↗

arXiv2018

AVA: A Video Dataset of Spatio-temporally Localized Atomic Visual Actions

This paper introduces a video dataset of spatio-temporally localized Atomic Visual Actions (AVA). The AVA dataset densely annotates 80 atomic visual actions in 430 15-minute video clips, where actions are localized in space and time, resulting in 1.58M action labels with multiple labels per person occurring frequently. The key characteristics of our dataset are: (1) the definition of atomic visual actions, rather than composite actions; (2) precise spatio-temporal annotations with possibly multiple annotations for each person; (3) exhaustive annotation of these atomic actions over 15-minute video clips; (4) people temporally linked across consecutive segments; and (5) using movies to gather a varied set of action representations. This departs from existing datasets for spatio-temporal action recognition, which typically provide sparse annotations for composite actions in short video clips. We will release the dataset publicly. AVA, with its realistic scene and action complexity, exposes the intrinsic difficulty of action recognition. To benchmark this, we present a novel approach for action localization that builds upon the current state-of-the-art methods, and demonstrates better performance on JHMDB and UCF101-24 categories. While setting a new state of the art on existing datasets, the overall results on AVA are low at 15.6% mAP, underscoring the need for developing new approaches for video understanding.

cs.CV↗

arXiv2018

The iNaturalist Species Classification and Detection Dataset

Existing image classification datasets used in computer vision tend to have a uniform distribution of images across object categories. In contrast, the natural world is heavily imbalanced, as some species are more abundant and easier to photograph than others. To encourage further progress in challenging real world conditions we present the iNaturalist species classification and detection dataset, consisting of 859,000 images from over 5,000 different species of plants and animals. It features visually similar species, captured in a wide variety of situations, from all over the world. Images were collected with different camera types, have varying image quality, feature a large class imbalance, and have been verified by multiple citizen scientists. We discuss the collection of the dataset and present extensive baseline experiments using state-of-the-art computer vision classification and detection models. Results show that current non-ensemble based methods achieve only 67% top one classification accuracy, illustrating the difficulty of the dataset. Specifically, we observe poor results for classes with small numbers of training examples suggesting more attention is needed in low-shot learning.

cs.CV↗

arXiv2018

A new solar neutrino channel for grand-unification monopole searches

We identify a previously untapped discovery channel for grand-unification monopoles, arising from their ability to catalyse the direct decay of protons into monoenergetic 459 MeV antineutrinos within the Sun. Previous analyses omit this possibility as it necessarily involves an electroweak suppression factor, and instead search for the unsuppressed 20-50 MeV neutrinos produced via two-stage proton decays. By accounting for the relative difference in interaction cross section and experimental background at typical neutrino detection experiments, we demonstrate that this new channel in fact possesses greater discovery potential. As a case in point, using 5326 live days of Super-Kamiokande exposure we find that $2\;σ$ ($3\;σ$) deviations in the 20-50 MeV channel are amplified to $3\;σ$ ($4.6\;σ$) deviations in the 459 MeV case. Exploiting correlations between these two channels may also offer even greater statistical power.

hep-ph↗

arXiv2018

Gravitational Radiation Background from Boson Star Binaries

We calculate the gravitational radiation background generated from boson star binaries formed in locally dense clusters with formation rate tracked by the regular star formation rate. We compute how the the frequency window in gravitational waves is affected by the boson field mass and repulsive self-coupling, anticipating constraints from EPTA and LISA. We also comment on the possible detectability of these binaries.

hep-ph↗