arXiv subjects

Shilong Liu

Publications and source records attributed to Shilong Liu.

98 records · Page 6Linked to original sources

arXiv2021

Unrestricted Adversarial Attacks on ImageNet Competition

Many works have investigated the adversarial attacks or defenses under the settings where a bounded and imperceptible perturbation can be added to the input. However in the real-world, the attacker does not need to comply with this restriction. In fact, more threats to the deep model come from unrestricted adversarial examples, that is, the attacker makes large and visible modifications on the image, which causes the model classifying mistakenly, but does not affect the normal observation in human perspective. Unrestricted adversarial attack is a popular and practical direction but has not been studied thoroughly. We organize this competition with the purpose of exploring more effective unrestricted adversarial attack algorithm, so as to accelerate the academical research on the model robustness under stronger unbounded attacks. The competition is held on the TianChi platform (\url{https://tianchi.aliyun.com/competition/entrance/531853/introduction}) as one of the series of AI Security Challengers Program.

cs.CV↗

arXiv2021

Query2Label: A Simple Transformer Way to Multi-Label Classification

This paper presents a simple and effective approach to solving the multi-label classification problem. The proposed approach leverages Transformer decoders to query the existence of a class label. The use of Transformer is rooted in the need of extracting local discriminative features adaptively for different labels, which is a strongly desired property due to the existence of multiple objects in one image. The built-in cross-attention module in the Transformer decoder offers an effective way to use label embeddings as queries to probe and pool class-related features from a feature map computed by a vision backbone for subsequent binary classifications. Compared with prior works, the new framework is simple, using standard Transformers and vision backbones, and effective, consistently outperforming all previous works on five multi-label classification data sets, including MS-COCO, PASCAL VOC, NUS-WIDE, and Visual Genome. Particularly, we establish $91.3\%$ mAP on MS-COCO. We hope its compact structure, simple implementation, and superior performance serve as a strong baseline for multi-label classification tasks and future studies. The code will be available soon at https://github.com/SlongLiu/query2labels.

cs.CV↗

arXiv2021

Unsupervised Part Segmentation through Disentangling Appearance and Shape

We study the problem of unsupervised discovery and segmentation of object parts, which, as an intermediate local representation, are capable of finding intrinsic object structure and providing more explainable recognition results. Recent unsupervised methods have greatly relaxed the dependency on annotated data which are costly to obtain, but still rely on additional information such as object segmentation mask or saliency map. To remove such a dependency and further improve the part segmentation performance, we develop a novel approach by disentangling the appearance and shape representations of object parts followed with reconstruction losses without using additional object mask information. To avoid degenerated solutions, a bottleneck block is designed to squeeze and expand the appearance representation, leading to a more effective disentanglement between geometry and appearance. Combined with a self-supervised part classification loss and an improved geometry concentration constraint, we can segment more consistent parts with semantic meanings. Comprehensive experiments on a wide variety of objects such as face, bird, and PASCAL VOC objects demonstrate the effectiveness of the proposed method.

cs.CV↗

arXiv2020

Increasing two-photon entangled dimensions by shaping input beam profiles

Photon pair entangled in high dimensional orbital angular momentum (OAM) degree of freedom (DOF) has been widely regarded as a possible source in improving the capacity of quantum information processing. The need for the generation of a high dimensional maximally entangled state in the OAM DOF is therefore much desired. In this work, we demonstrate a simple method to generate a broader and flatter OAM spectrum, i.e. a larger spiral bandwidth (SB), of entangled photon pairs generated through spontaneous parametric down-conversion by modifying the pump beam profile. By investigating both experimentally and theoretically, we have found that an exponential pump profile that is roughly the inverse of the mode profiles of the single-mode fibers used for OAM detection will provide a much larger SB when compared to a Gaussian shaped pump.

quant-ph↗

arXiv2019

A high-dimensional quantum frequency converter

In high dimensional quantum communication networks, quantum frequency convertor (QFC) is indispensable as an interface in the frequency domain. For example, many QFCs have been built to link atomic memories and fiber channels. However, almost all of QFCs work in a two-dimensional space. It is still a pivotal challenge to construct a high-quality QFC for some complex quantum states, e.g., a high dimensional single-photon state that refers to a qudit. Here, we firstly propose a high-dimensional QFC for an orbital angular momentum qudit via sum frequency conversion with a flat top beam pump. As a proof-of-principle demonstration, we realize quantum frequency conversions for a qudit from infrared to visible range. Based on the qudit quantum state tomography, the fidelities of converted state are 98.29(95.02)\%, 97.42(91.74)\%, and 86.75(67.04)\% for a qudit without (with) dark counts in 2,3, and 5 dimensions, respectively. The demonstration is very promising for constructing a high capacity quantum communication network.

quant-ph↗

arXiv2019

Multiplexing heralded single-photon in orbital angular momentum space

Heralded single-photon source (HSPS) with competitive single photon purity and indistinguishability has become an essential resource for photonic quantum information processing. Here, for the first time, we proposed a theoretical regime to enhance heralded single-photons generation by multiplexing the degree of the freedom of orbital angular momentum (OAM) of down-converted entangled photon pairs emitted from a nonlinear crystal. Experimentally, a proof-of-principle experiment has been performed through multiplexing three OAM modes. We achieve a 47$\%$ enhancement in single photon rate. A second-order autocorrelation function $g^{(2)}(0)<0.5$ ensures our multiplexed heralded single photons with good single photon purity. We further indicate that an OAM-multiplexed HSPS with high quality can be constructed by generating higher dimensional entangled state and sorting them with high efficiency in OAM space. Our avenue may approach a good HSPS with the deterministic property.

quant-ph↗

arXiv2019

Classical simulation of high-dimensional entanglement by non-separable angular--radial modes

An analogous model system for high-dimensional quantum entanglement is proposed, based on the angular and radial degrees of freedom of the improved Laguerre Gaussian mode. Experimentally, we observed strong violations of the Bell-CGLMP inequality for maximally non-separable states of dimension 2 through 10. The results for violations in classical non-separable state are in very good agreement with quantum instance, which illustrates that our scheme can be a useful platform to simulate high-dimensional non-local entanglement. Additionally, we found that the Bell measurements provide sufficient criteria for identifying mode separability in a high-dimensional space. Similar to the two-dimensional spin-orbit non-separable state, the proposed high-dimensional angular-radial non-separable state may provide promising applications for classical and quantum information processing.

quant-ph↗

arXiv2017

Efficient 525nm laser generation in single or double resonant cavity

This paper reports the results of a study into highly efficient sum frequency generation from 792 and 1556 nm wavelength light to 525 nm wavelength light using either a single or double resonant ring cavity based on a periodically poled potassium titanyl phosphate crystal (PPKTP). By optimizing the cavity$'$s parameters, the maximum power achieved for the resultant 525 nm laser was 263 and 373 mW for the single and double resonant cavity, respectively. The corresponding quantum conversion efficiencies were 8 and 77\% for converting 1556 nm photons to 525 nm photons with the single and double resonant cavity, respectively. The measured intra-cavity single pass conversion efficiency for both configurations was about 5\%. The performances of the sum frequency generation in these two configurations were studied and compared in detail. This work will provide guidelines for optimizing the generation of sum frequency generated laser light for a variety of configurations. The high conversion efficiency achieved in this work will help pave the way for frequency up-conversion of non-classical quantum states, such as the squeezed vacuum and single photon states. The proposed green laser source will be used in our future experiments, which includes a plan to generate two-colour entangled photon pairs and achieve the frequency down-conversion of single photons carrying orbital angular momentum.

physics.optics↗