arXiv subjects

Haifeng Liu

Publications and source records attributed to Haifeng Liu.

At least 37 records · Page 2Linked to original sources

arXiv2024

Semi-supervised 3D Object Detection with PatchTeacher and PillarMix

Semi-supervised learning aims to leverage numerous unlabeled data to improve the model performance. Current semi-supervised 3D object detection methods typically use a teacher to generate pseudo labels for a student, and the quality of the pseudo labels is essential for the final performance. In this paper, we propose PatchTeacher, which focuses on partial scene 3D object detection to provide high-quality pseudo labels for the student. Specifically, we divide a complete scene into a series of patches and feed them to our PatchTeacher sequentially. PatchTeacher leverages the low memory consumption advantage of partial scene detection to process point clouds with a high-resolution voxelization, which can minimize the information loss of quantization and extract more fine-grained features. However, it is non-trivial to train a detector on fractions of the scene. Therefore, we introduce three key techniques, i.e., Patch Normalizer, Quadrant Align, and Fovea Selection, to improve the performance of PatchTeacher. Moreover, we devise PillarMix, a strong data augmentation strategy that mixes truncated pillars from different LiDAR scans to generate diverse training samples and thus help the model learn more general representation. Extensive experiments conducted on Waymo and ONCE datasets verify the effectiveness and superiority of our method and we achieve new state-of-the-art results, surpassing existing methods by a large margin. Codes are available at https://github.com/LittlePey/PTPM.

cs.CV↗

arXiv2024

Less is More: Understanding Word-level Textual Adversarial Attack via n-gram Frequency Descend

Word-level textual adversarial attacks have demonstrated notable efficacy in misleading Natural Language Processing (NLP) models. Despite their success, the underlying reasons for their effectiveness and the fundamental characteristics of adversarial examples (AEs) remain obscure. This work aims to interpret word-level attacks by examining their $n$-gram frequency patterns. Our comprehensive experiments reveal that in approximately 90\% of cases, word-level attacks lead to the generation of examples where the frequency of $n$-grams decreases, a tendency we term as the $n$-gram Frequency Descend ($n$-FD). This finding suggests a straightforward strategy to enhance model robustness: training models using examples with $n$-FD. To examine the feasibility of this strategy, we employed the $n$-gram frequency information, as an alternative to conventional loss gradients, to generate perturbed examples in adversarial training. The experiment results indicate that the frequency-based approach performs comparably with the gradient-based approach in improving model robustness. Our research offers a novel and more intuitive perspective for understanding word-level textual adversarial attacks and proposes a new direction to improve model robustness.

cs.CL↗

arXiv2024

Boosting Few-Shot Learning via Attentive Feature Regularization

Few-shot learning (FSL) based on manifold regularization aims to improve the recognition capacity of novel objects with limited training samples by mixing two samples from different categories with a blending factor. However, this mixing operation weakens the feature representation due to the linear interpolation and the overlooking of the importance of specific channels. To solve these issues, this paper proposes attentive feature regularization (AFR) which aims to improve the feature representativeness and discriminability. In our approach, we first calculate the relations between different categories of semantic labels to pick out the related features used for regularization. Then, we design two attention-based calculations at both the instance and channel levels. These calculations enable the regularization procedure to focus on two crucial aspects: the feature complementarity through adaptive interpolation in related categories and the emphasis on specific feature channels. Finally, we combine these regularization strategies to significantly improve the classifier performance. Empirical studies on several popular FSL benchmarks demonstrate the effectiveness of AFR, which improves the recognition accuracy of novel categories without the need to retrain any feature extractor, especially in the 1-shot setting. Furthermore, the proposed AFR can seamlessly integrate into other FSL methods to improve classification performance.

cs.CV↗

arXiv2024

A Knowledge-Injected Curriculum Pretraining Framework for Question Answering

Knowledge-based question answering (KBQA) is a key task in NLP research, and also an approach to access the web data and knowledge, which requires exploiting knowledge graphs (KGs) for reasoning. In the literature, one promising solution for KBQA is to incorporate the pretrained language model (LM) with KGs by generating KG-centered pretraining corpus, which has shown its superiority. However, these methods often depend on specific techniques and resources to work, which may not always be available and restrict its application. Moreover, existing methods focus more on improving language understanding with KGs, while neglect the more important human-like complex reasoning. To this end, in this paper, we propose a general Knowledge-Injected Curriculum Pretraining framework (KICP) to achieve comprehensive KG learning and exploitation for KBQA tasks, which is composed of knowledge injection (KI), knowledge adaptation (KA) and curriculum reasoning (CR). Specifically, the KI module first injects knowledge into the LM by generating KG-centered pretraining corpus, and generalizes the process into three key steps that could work with different implementations for flexible application. Next, the KA module learns knowledge from the generated corpus with LM equipped with an adapter as well as keeps its original natural language understanding ability to reduce the negative impacts of the difference between the generated and natural corpus. Last, to enable the LM with complex reasoning, the CR module follows human reasoning patterns to construct three corpora with increasing difficulties of reasoning, and further trains the LM from easy to hard in a curriculum manner. We provide an implementation of the general framework, and evaluate the proposed KICP on four real-word datasets. The results demonstrate that our framework can achieve higher performances.

cs.CL↗

arXiv2023

TagCLIP: A Local-to-Global Framework to Enhance Open-Vocabulary Multi-Label Classification of CLIP Without Training

Contrastive Language-Image Pre-training (CLIP) has demonstrated impressive capabilities in open-vocabulary classification. The class token in the image encoder is trained to capture the global features to distinguish different text descriptions supervised by contrastive loss, making it highly effective for single-label classification. However, it shows poor performance on multi-label datasets because the global feature tends to be dominated by the most prominent class and the contrastive nature of softmax operation aggravates it. In this study, we observe that the multi-label classification results heavily rely on discriminative local features but are overlooked by CLIP. As a result, we dissect the preservation of patch-wise spatial information in CLIP and proposed a local-to-global framework to obtain image tags. It comprises three steps: (1) patch-level classification to obtain coarse scores; (2) dual-masking attention refinement (DMAR) module to refine the coarse scores; (3) class-wise reidentification (CWR) module to remedy predictions from a global perspective. This framework is solely based on frozen CLIP and significantly enhances its multi-label classification performance on various benchmarks without dataset-specific training. Besides, to comprehensively assess the quality and practicality of generated tags, we extend their application to the downstream task, i.e., weakly supervised semantic segmentation (WSSS) with generated tags as image-level pseudo labels. Experiments demonstrate that this classify-then-segment paradigm dramatically outperforms other annotation-free segmentation methods and validates the effectiveness of generated tags. Our code is available at https://github.com/linyq2117/TagCLIP.

cs.CV↗

arXiv2023

Heroes: Lightweight Federated Learning with Neural Composition and Adaptive Local Update in Heterogeneous Edge Networks

Federated Learning (FL) enables distributed clients to collaboratively train models without exposing their private data. However, it is difficult to implement efficient FL due to limited resources. Most existing works compress the transmitted gradients or prune the global model to reduce the resource cost, but leave the compressed or pruned parameters under-optimized, which degrades the training performance. To address this issue, the neural composition technique constructs size-adjustable models by composing low-rank tensors, allowing every parameter in the global model to learn the knowledge from all clients. Nevertheless, some tensors can only be optimized by a small fraction of clients, thus the global model may get insufficient training, leading to a long completion time, especially in heterogeneous edge scenarios. To this end, we enhance the neural composition technique, enabling all parameters to be fully trained. Further, we propose a lightweight FL framework, called Heroes, with enhanced neural composition and adaptive local update. A greedy-based algorithm is designed to adaptively assign the proper tensors and local update frequencies for participating clients according to their heterogeneous capabilities and resource budgets. Extensive experiments demonstrate that Heroes can reduce traffic consumption by about 72.05\% and provide up to 2.97$\times$ speedup compared to the baselines.

cs.DC↗

arXiv2023

Reducing Idleness in Financial Cloud Services via Multi-objective Evolutionary Reinforcement Learning based Load Balancer

In recent years, various companies have started to shift their data services from traditional data centers to the cloud. One of the major motivations is to save on operational costs with the aid of cloud elasticity. This paper discusses an emerging need from financial services to reduce the incidence of idle servers retaining very few user connections, without disconnecting them from the server side. This paper considers this need as a bi-objective online load balancing problem. A neural network based scalable policy is designed to route user requests to varied numbers of servers for the required elasticity. An evolutionary multi-objective training framework is proposed to optimize the weights of the policy. Not only is the new objective of idleness is reduced by over 130% more than traditional industrial solutions, but the original load balancing objective itself is also slightly improved. Extensive simulations with both synthetic and real-world data help reveal the detailed applicability of the proposed method to the emergent problem of reducing idleness in financial services.

cs.NE↗

arXiv2023

YOLO-based Semantic Communication with Generative AI-aided Resource Allocation for Digital Twins Construction

Digital Twins play a crucial role in bridging the physical and virtual worlds. Given the dynamic and evolving characteristics of the physical world, a huge volume of data transmission and exchange is necessary to attain synchronized updates in the virtual world. In this paper, we propose a semantic communication framework based on You Only Look Once (YOLO) to construct a virtual apple orchard with the aim of mitigating the costs associated with data transmission. Specifically, we first employ the YOLOv7-X object detector to extract semantic information from captured images of edge devices, thereby reducing the volume of transmitted data and saving transmission costs. Afterwards, we quantify the importance of each semantic information by the confidence generated through the object detector. Based on this, we propose two resource allocation schemes, i.e., the confidence-based scheme and the artificial intelligence-generated scheme, aimed at enhancing the transmission quality of important semantic information. The proposed diffusion model generates an optimal allocation scheme that outperforms both the average allocation scheme and the confidence-based allocation scheme. Moreover, to obtain semantic information more effectively, we enhance the detection capability of the YOLOv7-X object detector by introducing new Efficient Layer Aggregation Network-HorNet (ELAN-H) and SimAM attention modules, while reducing the model parameters and computational complexity, making it easier to run on edge devices with limited performance. The numerical results indicate that our proposed semantic communication framework and resource allocation schemes significantly reduce transmission costs while enhancing the transmission quality of important information in communication services.

cs.NI↗

arXiv2023

Random-walk shielding-potential viscosity model for warm dense metals

We develop a novel model, called the ``random-walk shielding-potential viscosity model'' (RWSP-VM) that introduces the statistics of random-walk ions and the Debye shielding effect to describe the viscosities of warm dense metals. The viscosities of several metals with low to high atomic number (Be, Al, Fe, and U) are calculated using the analytical expression of RWSP-VM. Additionally, we simulate the viscosities of Fe and Be by employing the Langevin molecular dynamics (MD) and classical MD, while the MD data for Al and U are obtained from a previous work. The results of the RWSP-VM are in good agreement with the MD results, which validates the proposed model. Furthermore, we compare the RWSP-VM with the one-component plasma model and Yukawa viscosity model and show that the three models yield results in excellent agreement with each other in the regime where the RWSP-VM is applicable. These results indicate that the RWSP-VM is a universal, accurate, and highly efficient model for calculating the viscosity of metals in the warm dense state. The code of the proposed RWSP-VM is provided, and it is envisaged that it will have broad application prospects in numerous fields.

cond-mat.stat-mech↗

arXiv2023

CLIP is Also an Efficient Segmenter: A Text-Driven Approach for Weakly Supervised Semantic Segmentation

Weakly supervised semantic segmentation (WSSS) with image-level labels is a challenging task. Mainstream approaches follow a multi-stage framework and suffer from high training costs. In this paper, we explore the potential of Contrastive Language-Image Pre-training models (CLIP) to localize different categories with only image-level labels and without further training. To efficiently generate high-quality segmentation masks from CLIP, we propose a novel WSSS framework called CLIP-ES. Our framework improves all three stages of WSSS with special designs for CLIP: 1) We introduce the softmax function into GradCAM and exploit the zero-shot ability of CLIP to suppress the confusion caused by non-target classes and backgrounds. Meanwhile, to take full advantage of CLIP, we re-explore text inputs under the WSSS setting and customize two text-driven strategies: sharpness-based prompt selection and synonym fusion. 2) To simplify the stage of CAM refinement, we propose a real-time class-aware attention-based affinity (CAA) module based on the inherent multi-head self-attention (MHSA) in CLIP-ViTs. 3) When training the final segmentation model with the masks generated by CLIP, we introduced a confidence-guided loss (CGL) focus on confident regions. Our CLIP-ES achieves SOTA performance on Pascal VOC 2012 and MS COCO 2014 while only taking 10% time of previous methods for the pseudo mask generation. Code is available at https://github.com/linyq2117/CLIP-ES.

cs.CV↗

arXiv2023

Data-Driven Prediction of Complex Crystal Structures of Dense Lithium

Lithium (Li) is a prototypical simple metal at ambient conditions, but exhibits remarkable changes in structural and electronic properties under compression. There has been intense debate about the structure of dense Li, and recent experiments offered fresh evidence for new yet undetermined crystalline phases near the enigmatic melting minimum region in the pressure-temperature phase diagram of Li. Here, we report on an extensive exploration of the energy landscape of Li using an advanced crystal structure search method combined with a machine learning approach, which greatly expands the scale of structure search, leading to the prediction of four complex Li crystal phases containing up to 192 atoms in the unit cell that are energetically competitive with known Li structures. These findings provide a viable solution to the newly observed yet unidentified crystalline phases of Li, and showcase the predictive power of the global structure search method for discovering complex crystal structures in conjunction with accurate machine-learning potentials.

cond-mat.mtrl-sci↗

arXiv2022

Boosting Semi-Supervised 3D Object Detection with Semi-Sampling

Current 3D object detection methods heavily rely on an enormous amount of annotations. Semi-supervised learning can be used to alleviate this issue. Previous semi-supervised 3D object detection methods directly follow the practice of fully-supervised methods to augment labeled and unlabeled data, which is sub-optimal. In this paper, we design a data augmentation method for semi-supervised learning, which we call Semi-Sampling. Specifically, we use ground truth labels and pseudo labels to crop gt samples and pseudo samples on labeled frames and unlabeled frames, respectively. Then we can generate a gt sample database and a pseudo sample database. When training a teacher-student semi-supervised framework, we randomly select gt samples and pseudo samples to both labeled frames and unlabeled frames, making a strong data augmentation for them. Our semi-sampling can be regarded as an extension of gt-sampling to semi-supervised learning. Our method is simple but effective. We consistently improve state-of-the-art methods on ScanNet, SUN-RGBD, and KITTI benchmarks by large margins. For example, when training using only 10% labeled data on ScanNet, we achieve 3.1 mAP and 6.4 mAP improvement upon 3DIoUMatch in terms of mAP@0.25 and mAP@0.5. When training using only 1% labeled data on KITTI, we boost 3DIoUMatch by 3.5 mAP, 6.7 mAP and 14.1 mAP on car, pedestrian and cyclist classes. Codes will be made publicly available at https://github.com/LittlePey/Semi-Sampling.

cs.CV↗

arXiv2022

Convolutional Embedding Makes Hierarchical Vision Transformer Stronger

Vision Transformers (ViTs) have recently dominated a range of computer vision tasks, yet it suffers from low training data efficiency and inferior local semantic representation capability without appropriate inductive bias. Convolutional neural networks (CNNs) inherently capture regional-aware semantics, inspiring researchers to introduce CNNs back into the architecture of the ViTs to provide desirable inductive bias for ViTs. However, is the locality achieved by the micro-level CNNs embedded in ViTs good enough? In this paper, we investigate the problem by profoundly exploring how the macro architecture of the hybrid CNNs/ViTs enhances the performances of hierarchical ViTs. Particularly, we study the role of token embedding layers, alias convolutional embedding (CE), and systemically reveal how CE injects desirable inductive bias in ViTs. Besides, we apply the optimal CE configuration to 4 recently released state-of-the-art ViTs, effectively boosting the corresponding performances. Finally, a family of efficient hybrid CNNs/ViTs, dubbed CETNets, are released, which may serve as generic vision backbones. Specifically, CETNets achieve 84.9% Top-1 accuracy on ImageNet-1K (training from scratch), 48.6% box mAP on the COCO benchmark, and 51.6% mIoU on the ADE20K, substantially improving the performances of the corresponding state-of-the-art baselines.

cs.CV↗

arXiv2022

Lidar Point Cloud Guided Monocular 3D Object Detection

Monocular 3D object detection is a challenging task in the self-driving and computer vision community. As a common practice, most previous works use manually annotated 3D box labels, where the annotating process is expensive. In this paper, we find that the precisely and carefully annotated labels may be unnecessary in monocular 3D detection, which is an interesting and counterintuitive finding. Using rough labels that are randomly disturbed, the detector can achieve very close accuracy compared to the one using the ground-truth labels. We delve into this underlying mechanism and then empirically find that: concerning the label accuracy, the 3D location part in the label is preferred compared to other parts of labels. Motivated by the conclusions above and considering the precise LiDAR 3D measurement, we propose a simple and effective framework, dubbed LiDAR point cloud guided monocular 3D object detection (LPCG). This framework is capable of either reducing the annotation costs or considerably boosting the detection accuracy without introducing extra annotation costs. Specifically, It generates pseudo labels from unlabeled LiDAR point clouds. Thanks to accurate LiDAR 3D measurements in 3D space, such pseudo labels can replace manually annotated labels in the training of monocular 3D detectors, since their 3D location information is precise. LPCG can be applied into any monocular 3D detector to fully use massive unlabeled data in a self-driving system. As a result, in KITTI benchmark, we take the first place on both monocular 3D and BEV (bird's-eye-view) detection with a significant margin. In Waymo benchmark, our method using 10% labeled data achieves comparable accuracy to the baseline detector using 100% labeled data. The codes are released at https://github.com/SPengLiang/LPCG.

cs.CV↗

arXiv2022

DID-M3D: Decoupling Instance Depth for Monocular 3D Object Detection

Monocular 3D detection has drawn much attention from the community due to its low cost and setup simplicity. It takes an RGB image as input and predicts 3D boxes in the 3D space. The most challenging sub-task lies in the instance depth estimation. Previous works usually use a direct estimation method. However, in this paper we point out that the instance depth on the RGB image is non-intuitive. It is coupled by visual depth clues and instance attribute clues, making it hard to be directly learned in the network. Therefore, we propose to reformulate the instance depth to the combination of the instance visual surface depth (visual depth) and the instance attribute depth (attribute depth). The visual depth is related to objects' appearances and positions on the image. By contrast, the attribute depth relies on objects' inherent attributes, which are invariant to the object affine transformation on the image. Correspondingly, we decouple the 3D location uncertainty into visual depth uncertainty and attribute depth uncertainty. By combining different types of depths and associated uncertainties, we can obtain the final instance depth. Furthermore, data augmentation in monocular 3D detection is usually limited due to the physical nature, hindering the boost of performance. Based on the proposed instance depth disentanglement strategy, we can alleviate this problem. Evaluated on KITTI, our method achieves new state-of-the-art results, and extensive ablation studies validate the effectiveness of each component in our method. The codes are released at https://github.com/SPengLiang/DID-M3D.

cs.CV↗

arXiv2022

Sparse Fuse Dense: Towards High Quality 3D Detection with Depth Completion

Current LiDAR-only 3D detection methods inevitably suffer from the sparsity of point clouds. Many multi-modal methods are proposed to alleviate this issue, while different representations of images and point clouds make it difficult to fuse them, resulting in suboptimal performance. In this paper, we present a novel multi-modal framework SFD (Sparse Fuse Dense), which utilizes pseudo point clouds generated from depth completion to tackle the issues mentioned above. Different from prior works, we propose a new RoI fusion strategy 3D-GAF (3D Grid-wise Attentive Fusion) to make fuller use of information from different types of point clouds. Specifically, 3D-GAF fuses 3D RoI features from the couple of point clouds in a grid-wise attentive way, which is more fine-grained and more precise. In addition, we propose a SynAugment (Synchronized Augmentation) to enable our multi-modal framework to utilize all data augmentation approaches tailored to LiDAR-only methods. Lastly, we customize an effective and efficient feature extractor CPConv (Color Point Convolution) for pseudo point clouds. It can explore 2D image features and 3D geometric features of pseudo point clouds simultaneously. Our method holds the highest entry on the KITTI car 3D object detection leaderboard, demonstrating the effectiveness of our SFD. Codes are available at https://github.com/LittlePey/SFD.

cs.CV↗

arXiv2022

Novel Valence Transition in Elemental Metal Europium around 80 GPa

Valence transition could induce structural, insulator-metal, nonmagnetic-magnetic and superconducting transitions in rare-earth metals and compounds, while the underlying physics remains unclear due to the complex interaction of localized 4f electrons as well as their coupling with itinerant electrons. The valence transition in the elemental metal europium (Eu) still has remained as a matter of debate. Using resonant x-ray emission scattering and x-ray diffraction, we pressurize the states of 4f electrons in Eu and study its valence and structure transitions up to 160 GPa. We provide compelling evidence for a valence transition around 80 GPa, which coincides with a structural transition from a monoclinic (C2/c) to an orthorhombic phase (Pnma). We show that the valence transition occurs when the pressure-dependent energy gap between 4f and 5d electrons approaches the Coulomb interaction. Our discovery is critical for understanding the electrodynamics of Eu, including magnetism and high-pressure superconductivity.

cond-mat.str-el↗

arXiv2022

Melting curve of magnesium up to 460 GPa from ab initio molecular dynamics simulations

Based on ab initio molecular dynamics simulations, we determined the melting curve of magnesium (Mg) up to ~460 GPa using the solid-liquid coexistence method. Between ~30 and 100 GPa, our melting curve is noticeably lower than those from static experiments, but is in good agreement with recent shock experiments. Up to ~450 GPa, our melting curve is generally consistent with the melting points from first-principles calculations using the small-cell coexistence method. We found that, at high pressures of a few hundred GPa, due to the strong softening of interatomic interactions in the liquid phase, solid-liquid coexistence simulations of Mg show some characteristics distinctively different from other metal systems, such as aluminum. For example, at a given volume, the pressure and temperature range for maintaining a stable solid-liquid coexistence state can be very small. The strong softening in the liquid phase also causes the unusual behavior of reentrant melting to occur at very high pressures. The onset of reentrant melting is predicted at ~305 GPa, close to that at ~300 GPa from the small-cell coexistence method. We show that the calculated melting points, considering reentrant melting, can be excellently fitted to a low-order Kechin equation, thereby making it possible for us to obtain a first-principles melting curve of Mg at pressures above 50 GPa for the first time. Similar characteristics in solid-liquid coexistence simulations, as well as reentrant melting, are also expected for other systems with strong softening in the liquid phase at high pressures.

cond-mat.mtrl-sci↗