arXiv subjects

Ran Cheng

Publications and source records attributed to Ran Cheng.

At least 127 records · Page 7Linked to original sources

arXiv2022

S3E-GNN: Sparse Spatial Scene Embedding with Graph Neural Networks for Camera Relocalization

Camera relocalization is the key component of simultaneous localization and mapping (SLAM) systems. This paper proposes a learning-based approach, named Sparse Spatial Scene Embedding with Graph Neural Networks (S3E-GNN), as an end-to-end framework for efficient and robust camera relocalization. S3E-GNN consists of two modules. In the encoding module, a trained S3E network encodes RGB images into embedding codes to implicitly represent spatial and semantic embedding code. With embedding codes and the associated poses obtained from a SLAM system, each image is represented as a graph node in a pose graph. In the GNN query module, the pose graph is transformed to form a embedding-aggregated reference graph for camera relocalization. We collect various scene datasets in the challenging environments to perform experiments. Our results demonstrate that S3E-GNN method outperforms the traditional Bag-of-words (BoW) for camera relocalization due to learning-based embedding and GNN powered scene matching mechanism.

cs.CV↗

arXiv2022

Theory of Harmonic Hall Responses of Spin-Torque Driven Antiferromagnets

Harmonic analysis is a powerful tool to characterize and quantify current-induced torques acting on magnetic materials, but so far it remains an open question in studying antiferromagnets. Here we formulate a general theory of harmonic Hall responses of collinear antiferromagnets driven by current-induced torques including both field-like and damping-like components. By scanning a magnetic field of variable strength in three orthogonal planes, we are able to distinguish the contributions from field-like torque, damping-like torque, and concomitant thermal effects by analyzing the second harmonic signals in the Hall voltage. The analytical expressions of the first and second harmonics as functions of the magnetic field direction and strength are confirmed by numerical simulations with good agreement. We demonstrate our predictions in two prototype antiferromagnets, $α-$Fe$_{2}$O$_{3}$ and NiO, providing direct and general guidance to current and future experiments.

cond-mat.mes-hall↗

arXiv2022

Semantic-Aware Pretraining for Dense Video Captioning

This report describes the details of our approach for the event dense-captioning task in ActivityNet Challenge 2021. We present a semantic-aware pretraining method for dense video captioning, which empowers the learned features to recognize high-level semantic concepts. Diverse video features of different modalities are fed into an event captioning module to generate accurate and meaningful sentences. Our final ensemble model achieves a 10.00 METEOR score on the test set.

cs.CV↗

arXiv2022

Ternary Compression for Communication-Efficient Federated Learning

Learning over massive data stored in different locations is essential in many real-world applications. However, sharing data is full of challenges due to the increasing demands of privacy and security with the growing use of smart mobile devices and IoT devices. Federated learning provides a potential solution to privacy-preserving and secure machine learning, by means of jointly training a global model without uploading data distributed on multiple devices to a central server. However, most existing work on federated learning adopts machine learning models with full-precision weights, and almost all these models contain a large number of redundant parameters that do not need to be transmitted to the server, consuming an excessive amount of communication costs. To address this issue, we propose a federated trained ternary quantization (FTTQ) algorithm, which optimizes the quantized networks on the clients through a self-learning quantization factor. Theoretical proofs of the convergence of quantization factors, unbiasedness of FTTQ, as well as a reduced weight divergence are given. On the basis of FTTQ, we propose a ternary federated averaging protocol (T-FedAvg) to reduce the upstream and downstream communication of federated learning systems. Empirical experiments are conducted to train widely used deep learning models on publicly available datasets, and our results demonstrate that the proposed T-FedAvg is effective in reducing communication costs and can even achieve slightly better performance on non-IID data in contrast to the canonical federated learning algorithms.

cs.LG↗

arXiv2022

A Perspective on Magnon Spin Nernst Effect in Antiferromagnets

Magnon excitations in antiferromagnetic materials and their physical implications enable novel device concepts not available in ferromagnets, emerging as a new area of active research. A unique characteristic of antiferromagnetic magnons is the coexistence of opposite spin polarization, which mimics the electron spin in a variety of transport phenomena. Among them, the most prominent spin-contrasting phenomenon is the magnon spin Nernst effect (SNE), which refers to the generation of transverse pure magnon spin current through a longitudinal temperature gradient. We introduce selected recent progress in the study of magnon SNE in collinear antiferromagnets with a focus on its underlying physical mechanism entailing profound topological features of the magnon band structures. By reviewing how the magnon SNE has inspired and enriched the exploration of topological magnons, we offer our perspectives on this emerging frontier that holds potential in future spintronic nano-technology.

cond-mat.mes-hall↗

arXiv2022

Manipulating Ferrimagnets by Fields and Currents

Ferrimagnets (FIMs) can function as high-frequency antiferromagnets while being easy to detect as ferromagnets, offering unique opportunities for ultrafast device applications. While the physical behavior of FIMs near the compensation point has been widely studied, there lacks a generic understanding of FIMs where the ratio of sublattice spins can vary freely between the ferromagnetic and antiferromagnetic limits. Here we investigate the physical properties of a model two-sublattice FIM manipulated by static magnetic fields and current-induced torques. By continuously varying the ratio of sublattice spins, we clarify how the dynamical chiral modes in an FIM are intrinsically connected to their ferro- and antiferromagnetic counterparts, which reveals unique features not visible near the compensation point. In particular, we find that current-induced torques can trigger spontaneous oscillation of the terahertz exchange mode. Compared with its realization in antiferromagnets, a spin-torque oscillator using FIMs not only has a reduced threshold current density but also can be self-stabilized, obviating the need for dynamic feedback.

cond-mat.mes-hall↗

arXiv2021

Quantifying Spin-Orbit Torques in Antiferromagnet/Heavy Metal Heterostructures

The effect of spin currents on the magnetic order of insulating antiferromagnets (AFMs) is of fundamental interest and can enable new applications. Toward this goal, characterizing the spin-orbit torques (SOT) associated with AFM/heavy metal (HM) interfaces is important. Here we report the full angular dependence of the harmonic Hall voltages in a predominantly easy-plane AFM, epitaxial c-axis oriented $α$-Fe$_2$O$_3$ films, with an interface to Pt. By modeling the harmonic Hall signals together with the $α$-Fe$_2$O$_3$ magnetic parameters, we determine the amplitudes of field-like and damping-like SOT. Out-of-plane field scans are shown to be essential to determining the damping-like component of the torques. In contrast to ferromagnetic/heavy metal heterostructures, our results demonstrate that the field-like torques are significantly larger than the damping-like torques, which we correlate with the presence of a large imaginary component of the interface spin-mixing conductance. Our work demonstrates a direct way of characterizing SOT in AFM/HM heterostructures.

cond-mat.mes-hall↗

arXiv2021

End-to-End Dense Video Captioning with Parallel Decoding

Dense video captioning aims to generate multiple associated captions with their temporal locations from the video. Previous methods follow a sophisticated "localize-then-describe" scheme, which heavily relies on numerous hand-crafted components. In this paper, we proposed a simple yet effective framework for end-to-end dense video captioning with parallel decoding (PDVC), by formulating the dense caption generation as a set prediction task. In practice, through stacking a newly proposed event counter on the top of a transformer decoder, the PDVC precisely segments the video into a number of event pieces under the holistic understanding of the video content, which effectively increases the coherence and readability of predicted captions. Compared with prior arts, the PDVC has several appealing advantages: (1) Without relying on heuristic non-maximum suppression or a recurrent event sequence selection network to remove redundancy, PDVC directly produces an event set with an appropriate size; (2) In contrast to adopting the two-stage scheme, we feed the enhanced representations of event queries into the localization head and caption head in parallel, making these two sub-tasks deeply interrelated and mutually promoted through the optimization; (3) Without bells and whistles, extensive experiments on ActivityNet Captions and YouCook2 show that PDVC is capable of producing high-quality captioning results, surpassing the state-of-the-art two-stage methods when its localization accuracy is on par with them. Code is available at https://github.com/ttengwang/PDVC.

cs.CV↗

arXiv2021

R4: A Framework for Route Representation and Route Recommendation

Route recommendation is significant in navigation service. Two major challenges for route recommendation are route representation and user representation. Different from items that can be identified by unique IDs in traditional recommendation, routes are combinations of links (i.e., a road segment and its following action like turning left) and the number of combinations could be close to infinite. Besides, the representation of a route changes under different scenarios. These facts result in severe sparsity of routes, which increases the difficulty of route representation. Moreover, link attribute deficiencies and errors affect preciseness of route representation. Because of the sparsity of routes, the interaction data between users and routes are also sparse. This makes it not easy to acquire user representation from historical user-item interactions as traditional recommendations do. To address these issues, we propose a novel learning framework R4. In R4, we design a sparse & dense network to obtain representations of routes. The sparse unit learns link ID embeddings and aggregates them to represent a route, which captures implicit route characteristics and subsequently alleviates problems caused by link attribute deficiencies and errors. The dense unit extracts implicit local features of routes from link attributes. For user representation, we utilize a series of historical navigation to extract user preference. R4 achieves remarkable performance in both offline and online experiments.

cs.AI↗

arXiv2021

Accelerating Multi-Objective Neural Architecture Search by Random-Weight Evaluation

For the goal of automated design of high-performance deep convolutional neural networks (CNNs), Neural Architecture Search (NAS) methodology is becoming increasingly important for both academia and industries.Due to the costly stochastic gradient descent (SGD) training of CNNs for performance evaluation, most existing NAS methods are computationally expensive for real-world deployments. To address this issue, we first introduce a new performance estimation metric, named Random-Weight Evaluation (RWE) to quantify the quality of CNNs in a cost-efficient manner. Instead of fully training the entire CNN, the RWE only trains its last layer and leaves the remainders with randomly initialized weights, which results in a single network evaluation in seconds.Second, a complexity metric is adopted for multi-objective NAS to balance the model size and performance. Overall, our proposed method obtains a set of efficient models with state-of-the-art performance in two real-world search spaces. Then the results obtained on the CIFAR-10 dataset are transferred to the ImageNet dataset to validate the practicality of the proposed algorithm. Moreover, ablation studies on NAS-Bench-301 datasets reveal the effectiveness of the proposed RWE in estimating the performance compared with existing methods.

cs.LG↗

arXiv2021

Revisiting Self-Training for Few-Shot Learning of Language Model

As unlabeled data carry rich task-relevant information, they are proven useful for few-shot learning of language model. The question is how to effectively make use of such data. In this work, we revisit the self-training technique for language model fine-tuning and present a state-of-the-art prompt-based few-shot learner, SFLM. Given two views of a text sample via weak and strong augmentation techniques, SFLM generates a pseudo label on the weakly augmented version. Then, the model predicts the same pseudo label when fine-tuned with the strongly augmented version. This simple approach is shown to outperform other state-of-the-art supervised and semi-supervised counterparts on six sentence classification and six sentence-pair classification benchmarking tasks. In addition, SFLM only relies on a few in-domain unlabeled data. We conduct a comprehensive analysis to demonstrate the robustness of our proposed approach under various settings, including augmentation techniques, model scale, and few-shot knowledge transfer across tasks.

cs.CL↗

arXiv2021

Spin Nernst Effect of Antiferromagnetic Magnons in the Presence of Spin Diffusion

Magnon spin Nernst effect was recently proposed as an intrinsic effect in antiferromagnets, where spin diffusion and boundary spin transmission have been ignored. However, diffusion processes are essential to convert a bulk spin current into boundary spin accumulation, which determines the spin injection rate into detectors through imperfect transmission. We formulate a diffusive theory to describe the detection of magnon spin Nernst effect with boundary conditions reflecting real device geometry. Thanks to the spin diffusion effect, the output signals in both electronic and optical detection grow rapidly with an increasing system size in the transverse dimension, which eventually saturate. Counterintuitively, the measurable signals are even functions of magnetic field, yielding optical detection more favorable than electronic detection.

cond-mat.mes-hall↗

arXiv2021

GP-S3Net: Graph-based Panoptic Sparse Semantic Segmentation Network

Panoptic segmentation as an integrated task of both static environmental understanding and dynamic object identification, has recently begun to receive broad research interest. In this paper, we propose a new computationally efficient LiDAR based panoptic segmentation framework, called GP-S3Net. GP-S3Net is a proposal-free approach in which no object proposals are needed to identify the objects in contrast to conventional two-stage panoptic systems, where a detection network is incorporated for capturing instance information. Our new design consists of a novel instance-level network to process the semantic results by constructing a graph convolutional network to identify objects (foreground), which later on are fused with the background classes. Through the fine-grained clusters of the foreground objects from the semantic segmentation backbone, over-segmentation priors are generated and subsequently processed by 3D sparse convolution to embed each cluster. Each cluster is treated as a node in the graph and its corresponding embedding is used as its node feature. Then a GCNN predicts whether edges exist between each cluster pair. We utilize the instance label to generate ground truth edge labels for each constructed graph in order to supervise the learning. Extensive experiments demonstrate that GP-S3Net outperforms the current state-of-the-art approaches, by a significant margin across available datasets such as, nuScenes and SemanticPOSS, ranking first on the competitive public SemanticKITTI leaderboard upon publication.

cs.CV↗

arXiv2021

FaPN: Feature-aligned Pyramid Network for Dense Image Prediction

Recent advancements in deep neural networks have made remarkable leap-forwards in dense image prediction. However, the issue of feature alignment remains as neglected by most existing approaches for simplicity. Direct pixel addition between upsampled and local features leads to feature maps with misaligned contexts that, in turn, translate to mis-classifications in prediction, especially on object boundaries. In this paper, we propose a feature alignment module that learns transformation offsets of pixels to contextually align upsampled higher-level features; and another feature selection module to emphasize the lower-level features with rich spatial details. We then integrate these two modules in a top-down pyramidal architecture and present the Feature-aligned Pyramid Network (FaPN). Extensive experimental evaluations on four dense prediction tasks and four datasets have demonstrated the efficacy of FaPN, yielding an overall improvement of 1.2 - 2.6 points in AP / mIoU over FPN when paired with Faster / Mask R-CNN. In particular, our FaPN achieves the state-of-the-art of 56.7% mIoU on ADE20K when integrated within Mask-Former. The code is available from https://github.com/EMI-Group/FaPN.

cs.CV↗

arXiv2021

RelativeNAS: Relative Neural Architecture Search via Slow-Fast Learning

Despite the remarkable successes of Convolutional Neural Networks (CNNs) in computer vision, it is time-consuming and error-prone to manually design a CNN. Among various Neural Architecture Search (NAS) methods that are motivated to automate designs of high-performance CNNs, the differentiable NAS and population-based NAS are attracting increasing interests due to their unique characters. To benefit from the merits while overcoming the deficiencies of both, this work proposes a novel NAS method, RelativeNAS. As the key to efficient search, RelativeNAS performs joint learning between fast-learners (i.e. networks with relatively higher accuracy) and slow-learners in a pairwise manner. Moreover, since RelativeNAS only requires low-fidelity performance estimation to distinguish each pair of fast-learner and slow-learner, it saves certain computation costs for training the candidate architectures. The proposed RelativeNAS brings several unique advantages: (1) it achieves state-of-the-art performance on ImageNet with top-1 error rate of 24.88%, i.e. outperforming DARTS and AmoebaNet-B by 1.82% and 1.12% respectively; (2) it spends only nine hours with a single 1080Ti GPU to obtain the discovered cells, i.e. 3.75x and 7875x faster than DARTS and AmoebaNet respectively; (3) it provides that the discovered cells obtained on CIFAR-10 can be directly transferred to object detection, semantic segmentation, and keypoint detection, yielding competitive results of 73.1% mAP on PASCAL VOC, 78.7% mIoU on Cityscapes, and 68.5% AP on MSCOCO, respectively. The implementation of RelativeNAS is available at https://github.com/EMI-Group/RelativeNAS

cs.CV↗

arXiv2021

Photoacoustic Silk Scaffolds for Neural stimulation and Regeneration

Neural interfaces using biocompatible scaffolds provide crucial properties for the functional repair of nerve injuries and neurodegenerative diseases, including cell adhesion, structural support, and mass transport. Neural stimulation has also been found to be effective in promoting neural regeneration. This work provides a new strategy to integrate photoacoustic (PA) neural stimulation into hydrogel scaffolds using a nanocomposite hydrogel approach. Specifically, polyethylene glycol (PEG)-functionalized carbon nanotubes (CNT), highly efficient photoacoustic agents, are embedded into silk fibroin to form biocompatible and soft photoacoustic materials. We show that these photoacoustic functional scaffolds enable non-genetic activation of neurons with a spatial precision defined by the area of light illumination, promoting neuron regeneration. These CNT/silk scaffolds offered reliable and repeatable photoacoustic neural stimulation. 94% of photoacoustic stimulated neurons exhibit a fluorescence change larger than 10% in calcium imaging in the light illuminated area. The on-demand photoacoustic stimulation increased neurite outgrowth by 1.74-fold in a dorsal root ganglion model, when compared to the unstimulated group. We also confirmed that photoacoustic neural stimulation promoted neurite outgrowth by impacting the brain-derived neurotrophic factor (BDNF) pathway. As a multifunctional neural scaffold, CNT/silk scaffolds demonstrated non-genetic PA neural stimulation functions and promoted neurite outgrowth, providing a new method for non-pharmacological neural regeneration.

q-bio.NC↗

arXiv2021

Spin Pumping of an Easy-Plane Antiferromagnet Enhanced by Dzyaloshinskii-Moriya Interaction

Recently, antiferromagnets have received revived interest due to their significant potential for developing next-generation ultrafast magnetic storage. Here we report dc spin pumping by the acoustic resonant mode in a canted easy-plane antiferromagnet α-Fe2O3 enabled by the Dzyaloshinskii-Moriya interaction. Systematic angle and frequency dependent measurements demonstrate that the observed spin pumping signals arise from resonance-induced spin injection and inverse spin Hall effect in α-Fe2O3/metal heterostructures, mimicking the behavior of spin pumping in conventional ferromagnet/nonmagnet systems. The pure spin current nature is further corroborated by reversal of the polarity of spin pumping signals when the spin detector is switched from platinum to tungsten which has an opposite sign of the spin Hall angle. Our results highlight the potential opportunities offered by the low-frequency acoustic resonant mode in canted easy-plane antiferromagnets for developing next-generation, functional spintronic devices.

cond-mat.mtrl-sci↗

arXiv2021

Lite-HDSeg: LiDAR Semantic Segmentation Using Lite Harmonic Dense Convolutions

Autonomous driving vehicles and robotic systems rely on accurate perception of their surroundings. Scene understanding is one of the crucial components of perception modules. Among all available sensors, LiDARs are one of the essential sensing modalities of autonomous driving systems due to their active sensing nature with high resolution of sensor readings. Accurate and fast semantic segmentation methods are needed to fully utilize LiDAR sensors for scene understanding. In this paper, we present Lite-HDSeg, a novel real-time convolutional neural network for semantic segmentation of full $3$D LiDAR point clouds. Lite-HDSeg can achieve the best accuracy vs. computational complexity trade-off in SemanticKitti benchmark and is designed on the basis of a new encoder-decoder architecture with light-weight harmonic dense convolutions as its core. Moreover, we introduce ICM, an improved global contextual module to capture multi-scale contextual features, and MCSPN, a multi-class Spatial Propagation Network to further refine the semantic boundaries. Our experimental results show that the proposed method outperforms state-of-the-art semantic segmentation approaches which can run real-time, thus is suitable for robotic and autonomous driving applications.

cs.CV↗