arXiv subjects

Haifeng Liu

Publications and source records attributed to Haifeng Liu.

At least 55 records · Page 3Linked to original sources

arXiv2022

Price DOES Matter! Modeling Price and Interest Preferences in Session-based Recommendation

Session-based recommendation aims to predict items that an anonymous user would like to purchase based on her short behavior sequence. The current approaches towards session-based recommendation only focus on modeling users' interest preferences, while they all ignore a key attribute of an item, i.e., the price. Many marketing studies have shown that the price factor significantly influences users' behaviors and the purchase decisions of users are determined by both price and interest preferences simultaneously. However, it is nontrivial to incorporate price preferences for session-based recommendation. Firstly, it is hard to handle heterogeneous information from various features of items to capture users' price preferences. Secondly, it is difficult to model the complex relations between price and interest preferences in determining user choices. To address the above challenges, we propose a novel method Co-guided Heterogeneous Hypergraph Network (CoHHN) for session-based recommendation. Towards the first challenge, we devise a heterogeneous hypergraph to represent heterogeneous information and rich relations among them. A dual-channel aggregating mechanism is then designed to aggregate various information in the heterogeneous hypergraph. After that, we extract users' price preferences and interest preferences via attention layers. As to the second challenge, a co-guided learning scheme is designed to model the relations between price and interest preferences and enhance the learning of each other. Finally, we predict user actions based on item features and users' price and interest preferences. Extensive experiments on three real-world datasets demonstrate the effectiveness of the proposed CoHHN. Further analysis reveals the significance of price for session-based recommendation.

cs.IR↗

arXiv2021

Suppress-and-Refine Framework for End-to-End 3D Object Detection

3D object detector based on Hough voting achieves great success and derives many follow-up works. Despite constantly refreshing the detection accuracy, these works suffer from handcrafted components used to eliminate redundant boxes, and thus are non-end-to-end and time-consuming. In this work, we propose a suppress-and-refine framework to remove these handcrafted components. To fully utilize full-resolution information and achieve real-time speed, it directly consumes feature points and redundant 3D proposals. Specifically, it first suppresses noisy 3D feature points and then feeds them to 3D proposals for the following RoI-aware refinement. With the gating mechanism to build fine proposal features and the self-attention mechanism to model relationships, our method can produce high-quality predictions with a small computation budget in an end-to-end manner. To this end, we present the first fully end-to-end 3D detector, SRDet, on the basis of VoteNet. It achieves state-of-the-art performance on the challenging ScanNetV2 and SUN RGB-D datasets with the fastest speed ever. Our code will be available at https://github.com/ZJULearning/SRDet.

cs.CV↗

arXiv2021

Class2Simi: A Noise Reduction Perspective on Learning with Noisy Labels

Learning with noisy labels has attracted a lot of attention in recent years, where the mainstream approaches are in pointwise manners. Meanwhile, pairwise manners have shown great potential in supervised metric learning and unsupervised contrastive learning. Thus, a natural question is raised: does learning in a pairwise manner mitigate label noise? To give an affirmative answer, in this paper, we propose a framework called Class2Simi: it transforms data points with noisy class labels to data pairs with noisy similarity labels, where a similarity label denotes whether a pair shares the class label or not. Through this transformation, the reduction of the noise rate is theoretically guaranteed, and hence it is in principle easier to handle noisy similarity labels. Amazingly, DNNs that predict the clean class labels can be trained from noisy data pairs if they are first pretrained from noisy data points. Class2Simi is computationally efficient because not only this transformation is on-the-fly in mini-batches, but also it just changes loss computation on top of model prediction into a pairwise manner. Its effectiveness is verified by extensive experiments.

cs.LG↗

arXiv2021

Accelerate CNNs from Three Dimensions: A Comprehensive Pruning Framework

Most neural network pruning methods, such as filter-level and layer-level prunings, prune the network model along one dimension (depth, width, or resolution) solely to meet a computational budget. However, such a pruning policy often leads to excessive reduction of that dimension, thus inducing a huge accuracy loss. To alleviate this issue, we argue that pruning should be conducted along three dimensions comprehensively. For this purpose, our pruning framework formulates pruning as an optimization problem. Specifically, it first casts the relationships between a certain model's accuracy and depth/width/resolution into a polynomial regression and then maximizes the polynomial to acquire the optimal values for the three dimensions. Finally, the model is pruned along the three optimal dimensions accordingly. In this framework, since collecting too much data for training the regression is very time-costly, we propose two approaches to lower the cost: 1) specializing the polynomial to ensure an accurate regression even with less training data; 2) employing iterative pruning and fine-tuning to collect the data faster. Extensive experiments show that our proposed algorithm surpasses state-of-the-art pruning algorithms and even neural architecture search-based algorithms.

cs.CV↗

arXiv2021

RESA: Recurrent Feature-Shift Aggregator for Lane Detection

Lane detection is one of the most important tasks in self-driving. Due to various complex scenarios (e.g., severe occlusion, ambiguous lanes, etc.) and the sparse supervisory signals inherent in lane annotations, lane detection task is still challenging. Thus, it is difficult for the ordinary convolutional neural network (CNN) to train in general scenes to catch subtle lane feature from the raw image. In this paper, we present a novel module named REcurrent Feature-Shift Aggregator (RESA) to enrich lane feature after preliminary feature extraction with an ordinary CNN. RESA takes advantage of strong shape priors of lanes and captures spatial relationships of pixels across rows and columns. It shifts sliced feature map recurrently in vertical and horizontal directions and enables each pixel to gather global information. RESA can conjecture lanes accurately in challenging scenarios with weak appearance clues by aggregating sliced feature map. Moreover, we propose a Bilateral Up-Sampling Decoder that combines coarse-grained and fine-detailed features in the up-sampling stage. It can recover the low-resolution feature map into pixel-wise prediction meticulously. Our method achieves state-of-the-art results on two popular lane detection benchmarks (CULane and Tusimple). Code has been made available at: https://github.com/ZJULearning/resa.

cs.CV↗

arXiv2021

Complementary Pseudo Labels For Unsupervised Domain Adaptation On Person Re-identification

In recent years, supervised person re-identification (re-ID) models have received increasing studies. However, these models trained on the source domain always suffer dramatic performance drop when tested on an unseen domain. Existing methods are primary to use pseudo labels to alleviate this problem. One of the most successful approaches predicts neighbors of each unlabeled image and then uses them to train the model. Although the predicted neighbors are credible, they always miss some hard positive samples, which may hinder the model from discovering important discriminative information of the unlabeled domain. In this paper, to complement these low recall neighbor pseudo labels, we propose a joint learning framework to learn better feature embeddings via high precision neighbor pseudo labels and high recall group pseudo labels. The group pseudo labels are generated by transitively merging neighbors of different samples into a group to achieve higher recall. However, the merging operation may cause subgroups in the group due to imperfect neighbor predictions. To utilize these group pseudo labels properly, we propose using a similarity-aggregating loss to mitigate the influence of these subgroups by pulling the input sample towards the most similar embeddings. Extensive experiments on three large-scale datasets demonstrate that our method can achieve state-of-the-art performance under the unsupervised domain adaptation re-ID setting.

cs.CV↗

arXiv2020

Part-dependent Label Noise: Towards Instance-dependent Label Noise

Learning with the \textit{instance-dependent} label noise is challenging, because it is hard to model such real-world noise. Note that there are psychological and physiological evidences showing that we humans perceive instances by decomposing them into parts. Annotators are therefore more likely to annotate instances based on the parts rather than the whole instances, where a wrong mapping from parts to classes may cause the instance-dependent label noise. Motivated by this human cognition, in this paper, we approximate the instance-dependent label noise by exploiting \textit{part-dependent} label noise. Specifically, since instances can be approximately reconstructed by a combination of parts, we approximate the instance-dependent \textit{transition matrix} for an instance by a combination of the transition matrices for the parts of the instance. The transition matrices for parts can be learned by exploiting anchor points (i.e., data points that belong to a specific class almost surely). Empirical evaluations on synthetic and real-world datasets demonstrate our method is superior to the state-of-the-art approaches for learning from the instance-dependent label noise.

cs.LG↗

arXiv2020

Socially-Aware Conference Participant Recommendation with Personality Traits

As a result of the importance of academic collaboration at smart conferences, various researchers have utilized recommender systems to generate effective recommendations for participants. Recent research has shown that the personality traits of users can be used as innovative entities for effective recommendations. Nevertheless, subjective perceptions involving the personality of participants at smart conferences are quite rare and haven't gained much attention. Inspired by the personality and social characteristics of users, we present an algorithm called Socially and Personality Aware Recommendation of Participants (SPARP). Our recommendation methodology hybridizes the computations of similar interpersonal relationships and personality traits among participants. SPARP models the personality and social characteristic profiles of participants at a smart conference. By combining the above recommendation entities, SPARP then recommends participants to each other for effective collaborations. We evaluate SPARP using a relevant dataset. Experimental results confirm that SPARP is reliable and outperforms other state-of-the-art methods.

cs.SI↗

arXiv2020

Scientific Article Recommendation: Exploiting Common Author Relations and Historical Preferences

Scientific article recommender systems are playing an increasingly important role for researchers in retrieving scientific articles of interest in the coming era of big scholarly data. Most existing studies have designed unified methods for all target researchers and hence the same algorithms are run to generate recommendations for all researchers no matter which situations they are in. However, different researchers may have their own features and there might be corresponding methods for them resulting in better recommendations. In this paper, we propose a novel recommendation method which incorporates information on common author relations between articles (i.e., two articles with the same author(s)). The rationale underlying our method is that researchers often search articles published by the same author(s). Since not all researchers have such author-based search patterns, we present two features, which are defined based on information about pairwise articles with common author relations and frequently appeared authors, to determine target researchers for recommendation. Extensive experiments we performed on a real-world dataset demonstrate that the defined features are effective to determine relevant target researchers and the proposed method generates more accurate recommendations for relevant researchers when compared to a Baseline method.

cs.SI↗

arXiv2020

Boundary-Aware Dense Feature Indicator for Single-Stage 3D Object Detection from Point Clouds

3D object detection based on point clouds has become more and more popular. Some methods propose localizing 3D objects directly from raw point clouds to avoid information loss. However, these methods come with complex structures and significant computational overhead, limiting its broader application in real-time scenarios. Some methods choose to transform the point cloud data into compact tensors first and leverage off-the-shelf 2D detectors to propose 3D objects, which is much faster and achieves state-of-the-art results. However, because of the inconsistency between 2D and 3D data, we argue that the performance of compact tensor-based 3D detectors is restricted if we use 2D detectors without corresponding modification. Specifically, the distribution of point clouds is uneven, with most points gather on the boundary of objects, while detectors for 2D data always extract features evenly. Motivated by this observation, we propose DENse Feature Indicator (DENFI), a universal module that helps 3D detectors focus on the densest region of the point clouds in a boundary-aware manner. Moreover, DENFI is lightweight and guarantees real-time speed when applied to 3D object detectors. Experiments on KITTI dataset show that DENFI improves the performance of the baseline single-stage detector remarkably, which achieves new state-of-the-art performance among previous 3D detectors, including both two-stage and multi-sensor fusion methods, in terms of mAP with a 34FPS detection speed.

cs.CV↗

arXiv2020

Multi-Class Classification from Noisy-Similarity-Labeled Data

A similarity label indicates whether two instances belong to the same class while a class label shows the class of the instance. Without class labels, a multi-class classifier could be learned from similarity-labeled pairwise data by meta classification learning. However, since the similarity label is less informative than the class label, it is more likely to be noisy. Deep neural networks can easily remember noisy data, leading to overfitting in classification. In this paper, we propose a method for learning from only noisy-similarity-labeled data. Specifically, to model the noise, we employ a noise transition matrix to bridge the class-posterior probability between clean and noisy data. We further estimate the transition matrix from only noisy data and build a novel learning system to learn a classifier which can assign noise-free class labels for instances. Moreover, we theoretically justify how our proposed method generalizes for learning classifiers. Experimental results demonstrate the superiority of the proposed method over the state-of-the-art method on benchmark-simulated and real-world noisy-label datasets.

cs.LG↗

arXiv2020

First-Principles Study of Hydrogen Behaviors in $α$-Pu$_{2}$O$_{3}$

The in-depth understanding of hydrogen permeation through plutonium-oxide overlayers is the prerequisite to evaluate the complex hydriding induction period of Pu. In this work, the incorporation, diffusion and dissolution of hydrogen in $α$-Pu$_{2}$O$_{3}$ are investigated by the first-principles calculations and $\textit{ab initio}$ thermodynamic method based on DFT+U and DFT-D3 schemes. Our study reveals that the hydrogen incorporation is endothermic and the separated H atoms prefer to recombine as H$_{2}$ molecules rather than reacting with $α$-Pu$_{2}$O$_{3}$. The H and H$_{2}$ diffusion are both feasible, generally, H will recombine first as H$_{2}$ and then migrate. Both pressure P$_{H2}$ and temperature can promote the hydrogen dissolution in $α$-Pu$_{2}$O$_{3}$. The single H$_{2}$ molecule incorporation and (H+H$_{2}$) mixed dissolution will successively appear when increasing P$_{H2}$. Compared to PuO$_{2}$, this work indicates that Pu sesquioxide is hardly reduced by hydrogen, but the porous $α$-Pu$_{2}$O$_{3}$ facilitates hydrogen transport in Pu oxide layers. We presents the microscopic picture of hydrogen behaviors in the defect-free $α$-Pu$_{2}$O$_{3}$, which could shed some light on the study of the hydriding induction period of Pu.

cond-mat.mtrl-sci↗

arXiv2020

Adversarial-Learned Loss for Domain Adaptation

Recently, remarkable progress has been made in learning transferable representation across domains. Previous works in domain adaptation are majorly based on two techniques: domain-adversarial learning and self-training. However, domain-adversarial learning only aligns feature distributions between domains but does not consider whether the target features are discriminative. On the other hand, self-training utilizes the model predictions to enhance the discrimination of target features, but it is unable to explicitly align domain distributions. In order to combine the strengths of these two methods, we propose a novel method called Adversarial-Learned Loss for Domain Adaptation (ALDA). We first analyze the pseudo-label method, a typical self-training method. Nevertheless, there is a gap between pseudo-labels and the ground truth, which can cause incorrect training. Thus we introduce the confusion matrix, which is learned through an adversarial manner in ALDA, to reduce the gap and align the feature distributions. Finally, a new loss function is auto-constructed from the learned confusion matrix, which serves as the loss for unlabeled target samples. Our ALDA outperforms state-of-the-art approaches in four standard domain adaptation datasets. Our code is available at https://github.com/ZJULearning/ALDA.

cs.CV↗

arXiv2019

DBP: Discrimination Based Block-Level Pruning for Deep Model Acceleration

Neural network pruning is one of the most popular methods of accelerating the inference of deep convolutional neural networks (CNNs). The dominant pruning methods, filter-level pruning methods, evaluate their performance through the reduction ratio of computations and deem that a higher reduction ratio of computations is equivalent to a higher acceleration ratio in terms of inference time. However, we argue that they are not equivalent if parallel computing is considered. Given that filter-level pruning only prunes filters in layers and computations in a layer usually run in parallel, most computations reduced by filter-level pruning usually run in parallel with the un-reduced ones. Thus, the acceleration ratio of filter-level pruning is limited. To get a higher acceleration ratio, it is better to prune redundant layers because computations of different layers cannot run in parallel. In this paper, we propose our Discrimination based Block-level Pruning method (DBP). Specifically, DBP takes a sequence of consecutive layers (e.g., Conv-BN-ReLu) as a block and removes redundant blocks according to the discrimination of their output features. As a result, DBP achieves a considerable acceleration ratio by reducing the depth of CNNs. Extensive experiments show that DBP has surpassed state-of-the-art filter-level pruning methods in both accuracy and acceleration ratio. Our code will be made available soon.

cs.CV↗

arXiv2019

Training-Time-Friendly Network for Real-Time Object Detection

Modern object detectors can rarely achieve short training time, fast inference speed, and high accuracy at the same time. To strike a balance among them, we propose the Training-Time-Friendly Network (TTFNet). In this work, we start with light-head, single-stage, and anchor-free designs, which enable fast inference speed. Then, we focus on shortening training time. We notice that encoding more training samples from annotated boxes plays a similar role as increasing batch size, which helps enlarge the learning rate and accelerate the training process. To this end, we introduce a novel approach using Gaussian kernels to encode training samples. Besides, we design the initiative sample weights for better information utilization. Experiments on MS COCO show that our TTFNet has great advantages in balancing training time, inference speed, and accuracy. It has reduced training time by more than seven times compared to previous real-time detectors while maintaining state-of-the-art performances. In addition, our super-fast version of TTFNet-18 and TTFNet-53 can outperform SSD300 and YOLOv3 by less than one-tenth of their training time, respectively. The code has been made available at \url{https://github.com/ZJULearning/ttfnet}.

cs.CV↗

arXiv2019

CFS: A Distributed File System for Large Scale Container Platforms

We propose CFS, a distributed file system for large scale container platforms. CFS supports both sequential and random file accesses with optimized storage for both large files and small files, and adopts different replication protocols for different write scenarios to improve the replication performance. It employs a metadata subsystem to store and distribute the file metadata across different storage nodes based on the memory usage. This metadata placement strategy avoids the need of data rebalancing during capacity expansion. CFS also provides POSIX-compliant APIs with relaxed semantics and metadata atomicity to improve the system performance. We performed a comprehensive comparison with Ceph, a widely-used distributed file system on container platforms. Our experimental results show that, in testing 7 commonly used metadata operations, CFS gives around 3 times performance boost on average. In addition, CFS exhibits better random-read/write performance in highly concurrent environments with multiple clients and processes.

cs.DC↗

arXiv2019

Identification of important error fields in stellarators using Hessian matrix method

Error fields are predominantly attributed to inevitable coil imperfections. Controlling error fields during coil fabrication and assembly is crucial for stellarators. Excessively tight coil tolerance increases time and cost, and, in part, led to the cancellation of NCSX and delay of W7-X. In this paper, we improve the recently proposed Hessian matrix method to rapidly identify important coil deviations. Two of the most common figures of merit, magnetic island size and quasi-symmetry, are analytically differentiated over coil parameters. By extracting the eigenvectors of the Hessian matrix, we can directly identify sensitive coil deviations in the order of the eigenvalues. The new method is applied to the upcoming CFQS configuration. Important perturbations that enlarge n/m=4/11 islands and deteriorate quasi-axisymmetry of the magnetic field are successfully determined. The results suggest each modular coil should have separate tolerance and some certain perturbation combinations will produce significant error fields. By relaxing unnecessary coil tolerance, this method will hopefully lead to a substantial reduction in time and cost.

physics.plasm-ph↗

arXiv2019

The Design and Implementation of a Real Time Visual Search System on JD E-commerce Platform

We present the design and implementation of a visual search system for real time image retrieval on JD.com, the world's third largest and China's largest e-commerce site. We demonstrate that our system can support real time visual search with hundreds of billions of product images at sub-second timescales and handle frequent image updates through distributed hierarchical architecture and efficient indexing methods. We hope that sharing our practice with our real production system will inspire the middleware community's interest and appreciation for building practical large scale systems for emerging applications, such as ecommerce visual search.

cs.IR↗