Searcharxiv⌕ Search

arXiv subjects

Feng Zhao

Publications and source records attributed to Feng Zhao.

At least 145 records · Page 8Linked to original sources
arXiv2024

Point-DETR3D: Leveraging Imagery Data with Spatial Point Prior for Weakly Semi-supervised 3D Object Detection

Training high-accuracy 3D detectors necessitates massive labeled 3D annotations with 7 degree-of-freedom, which is laborious and time-consuming. Therefore, the form of point annotations is proposed to offer significant prospects for practical applications in 3D detection, which is not only more accessible and less expensive but also provides strong spatial information for object localization. In this paper, we empirically discover that it is non-trivial to merely adapt Point-DETR to its 3D form, encountering two main bottlenecks: 1) it fails to encode strong 3D prior into the model, and 2) it generates low-quality pseudo labels in distant regions due to the extreme sparsity of LiDAR points. To overcome these challenges, we introduce Point-DETR3D, a teacher-student framework for weakly semi-supervised 3D detection, designed to fully capitalize on point-wise supervision within a constrained instance-wise annotation budget.Different from Point-DETR which encodes 3D positional information solely through a point encoder, we propose an explicit positional query initialization strategy to enhance the positional prior. Considering the low quality of pseudo labels at distant regions produced by the teacher model, we enhance the detector's perception by incorporating dense imagery data through a novel Cross-Modal Deformable RoI Fusion (D-RoI).Moreover, an innovative point-guided self-supervised learning technique is proposed to allow for fully exploiting point priors, even in student models.Extensive experiments on representative nuScenes dataset demonstrate our Point-DETR3D obtains significant improvements compared to previous works. Notably, with only 5% of labeled data, Point-DETR3D achieves over 90% performance of its fully supervised counterpart.

cs.CV↗
arXiv2024

Agent-FLAN: Designing Data and Methods of Effective Agent Tuning for Large Language Models

Open-sourced Large Language Models (LLMs) have achieved great success in various NLP tasks, however, they are still far inferior to API-based models when acting as agents. How to integrate agent ability into general LLMs becomes a crucial and urgent problem. This paper first delivers three key observations: (1) the current agent training corpus is entangled with both formats following and agent reasoning, which significantly shifts from the distribution of its pre-training data; (2) LLMs exhibit different learning speeds on the capabilities required by agent tasks; and (3) current approaches have side-effects when improving agent abilities by introducing hallucinations. Based on the above findings, we propose Agent-FLAN to effectively Fine-tune LANguage models for Agents. Through careful decomposition and redesign of the training corpus, Agent-FLAN enables Llama2-7B to outperform prior best works by 3.5\% across various agent evaluation datasets. With comprehensively constructed negative samples, Agent-FLAN greatly alleviates the hallucination issues based on our established evaluation benchmark. Besides, it consistently improves the agent capability of LLMs when scaling model sizes while slightly enhancing the general capability of LLMs. The code will be available at https://github.com/InternLM/Agent-FLAN.

cs.CL↗
arXiv2024

Decomposition Ascribed Synergistic Learning for Unified Image Restoration

Learning to restore multiple image degradations within a single model is quite beneficial for real-world applications. Nevertheless, existing works typically concentrate on regarding each degradation independently, while their relationship has been less exploited to ensure the synergistic learning. To this end, we revisit the diverse degradations through the lens of singular value decomposition, with the observation that the decomposed singular vectors and singular values naturally undertake the different types of degradation information, dividing various restoration tasks into two groups, \ie, singular vector dominated and singular value dominated. The above analysis renders a more unified perspective to ascribe the diverse degradations, compared to previous task-level independent learning. The dedicated optimization of degraded singular vectors and singular values inherently utilizes the potential relationship among diverse restoration tasks, attributing to the Decomposition Ascribed Synergistic Learning (DASL). Specifically, DASL comprises two effective operators, namely, Singular VEctor Operator (SVEO) and Singular VAlue Operator (SVAO), to favor the decomposed optimization, which can be lightly integrated into existing image restoration backbone. Moreover, the congruous decomposition loss has been devised for auxiliary. Extensive experiments on blended five image restoration tasks demonstrate the effectiveness of our method.

cs.CV↗
arXiv2024

Prompt Learning on Temporal Interaction Graphs

Temporal Interaction Graphs (TIGs) are widely utilized to represent real-world systems. To facilitate representation learning on TIGs, researchers have proposed a series of TIG models. However, these models are still facing two tough gaps between the pre-training and downstream predictions in their ``pre-train, predict'' training paradigm. First, the temporal discrepancy between the pre-training and inference data severely undermines the models' applicability in distant future predictions on the dynamically evolving data. Second, the semantic divergence between pretext and downstream tasks hinders their practical applications, as they struggle to align with their learning and prediction capabilities across application scenarios. Recently, the ``pre-train, prompt'' paradigm has emerged as a lightweight mechanism for model generalization. Applying this paradigm is a potential solution to solve the aforementioned challenges. However, the adaptation of this paradigm to TIGs is not straightforward. The application of prompting in static graph contexts falls short in temporal settings due to a lack of consideration for time-sensitive dynamics and a deficiency in expressive power. To address this issue, we introduce Temporal Interaction Graph Prompting (TIGPrompt), a versatile framework that seamlessly integrates with TIG models, bridging both the temporal and semantic gaps. In detail, we propose a temporal prompt generator to offer temporally-aware prompts for different tasks. These prompts stand out for their minimalistic design, relying solely on the tuning of the prompt generator with very little supervision data. To cater to varying computational resource demands, we propose an extended ``pre-train, prompt-based fine-tune'' paradigm, offering greater flexibility. Through extensive experiments, the TIGPrompt demonstrates the SOTA performance and remarkable efficiency advantages.

cs.AI↗
arXiv2024

A Vanilla Multi-Task Framework for Dense Visual Prediction Solution to 1st VCL Challenge -- Multi-Task Robustness Track

In this report, we present our solution to the multi-task robustness track of the 1st Visual Continual Learning (VCL) Challenge at ICCV 2023 Workshop. We propose a vanilla framework named UniNet that seamlessly combines various visual perception algorithms into a multi-task model. Specifically, we choose DETR3D, Mask2Former, and BinsFormer for 3D object detection, instance segmentation, and depth estimation tasks, respectively. The final submission is a single model with InternImage-L backbone, and achieves a 49.6 overall score (29.5 Det mAP, 80.3 mTPS, 46.4 Seg mAP, and 7.93 silog) on SHIFT validation set. Besides, we provide some interesting observations in our experiments which may facilitate the development of multi-task learning in dense visual prediction.

cs.CV↗
arXiv2024

On the expected number of facets for the convex hull of samples

This paper studies the convex hull of $d$-dimensional samples i.i.d. generated from spherically symmetric distributions. Specifically, we derive a complete integration formula for the expected facet number of the convex hull. This formula is with respect to the CDF of the radial distribution. As the number of samples approaches infinity, the integration formula enables us to obtain the asymptotic value of the expected facet number for three categories of spherically symmetric distributions. Additionally, the asymptotic result can be applied to estimating the sample complexity in order that the probability measure of the convex hull tends to one.

math.PR↗
arXiv2024

Stream Query Denoising for Vectorized HD Map Construction

To enhance perception performance in complex and extensive scenarios within the realm of autonomous driving, there has been a noteworthy focus on temporal modeling, with a particular emphasis on streaming methods. The prevailing trend in streaming models involves the utilization of stream queries for the propagation of temporal information. Despite the prevalence of this approach, the direct application of the streaming paradigm to the construction of vectorized high-definition maps (HD-maps) fails to fully harness the inherent potential of temporal information. This paper introduces the Stream Query Denoising (SQD) strategy as a novel approach for temporal modeling in high-definition map (HD-map) construction. SQD is designed to facilitate the learning of temporal consistency among map elements within the streaming model. The methodology involves denoising the queries that have been perturbed by the addition of noise to the ground-truth information from the preceding frame. This denoising process aims to reconstruct the ground-truth information for the current frame, thereby simulating the prediction process inherent in stream queries. The SQD strategy can be applied to those streaming methods (e.g., StreamMapNet) to enhance the temporal modeling. The proposed SQD-MapNet is the StreamMapNet equipped with SQD. Extensive experiments on nuScenes and Argoverse2 show that our method is remarkably superior to other existing methods across all settings of close range and long range. The code will be available soon.

cs.CV↗
arXiv2024

T-Eval: Evaluating the Tool Utilization Capability of Large Language Models Step by Step

Large language models (LLM) have achieved remarkable performance on various NLP tasks and are augmented by tools for broader applications. Yet, how to evaluate and analyze the tool-utilization capability of LLMs is still under-explored. In contrast to previous works that evaluate models holistically, we comprehensively decompose the tool utilization into multiple sub-processes, including instruction following, planning, reasoning, retrieval, understanding, and review. Based on that, we further introduce T-Eval to evaluate the tool utilization capability step by step. T-Eval disentangles the tool utilization evaluation into several sub-domains along model capabilities, facilitating the inner understanding of both holistic and isolated competency of LLMs. We conduct extensive experiments on T-Eval and in-depth analysis of various LLMs. T-Eval not only exhibits consistency with the outcome-oriented evaluation but also provides a more fine-grained analysis of the capabilities of LLMs, providing a new perspective in LLM evaluation on tool-utilization ability. The benchmark will be available at https://github.com/open-compass/T-Eval.

cs.CL↗
arXiv2023

ShareGPT4V: Improving Large Multi-Modal Models with Better Captions

In the realm of large multi-modal models (LMMs), efficient modality alignment is crucial yet often constrained by the scarcity of high-quality image-text data. To address this bottleneck, we introduce the ShareGPT4V dataset, a pioneering large-scale resource featuring 1.2 million highly descriptive captions, which surpasses existing datasets in diversity and information content, covering world knowledge, object properties, spatial relationships, and aesthetic evaluations. Specifically, ShareGPT4V originates from a curated 100K high-quality captions collected from advanced GPT4-Vision and has been expanded to 1.2M with a superb caption model trained on this subset. ShareGPT4V first demonstrates its effectiveness for the Supervised Fine-Tuning (SFT) phase, by substituting an equivalent quantity of detailed captions in existing SFT datasets with a subset of our high-quality captions, significantly enhancing the LMMs like LLaVA-7B, LLaVA-1.5-13B, and Qwen-VL-Chat-7B on the MME and MMBench benchmarks, with respective gains of 222.8/22.0/22.3 and 2.7/1.3/1.5. We further incorporate ShareGPT4V data into both the pre-training and SFT phases, obtaining ShareGPT4V-7B, a superior LMM based on a simple architecture that has remarkable performance across a majority of the multi-modal benchmarks. This project is available at https://ShareGPT4V.github.io to serve as a pivotal resource for advancing the LMMs community.

cs.CV↗
arXiv2023

Single-shot deterministic complex amplitude imaging with a single-layer metalens

Conventional imaging systems can only capture light intensity. Meanwhile, the lost phase information may be critical for a variety of applications such as label-free microscopy and optical metrology. Existing phase retrieval techniques typically require a bulky setup, multi-frame measurements, or prior information of the target scene. Here, we proposed an extremely compact system for complex amplitude imaging, leveraging the extreme versatility of a single-layer metalens to generate spatially-multiplexed and polarization-phase-shifted point spread functions. Combining the metalens with a polarization camera, the system can simultaneously record four polarization shearing interference patterns along both in-plane directions, thus allowing the deterministic reconstruction of the complex amplitude light field in a single shot. Using an incoherent light-emitting diode as the illumination, we experimentally demonstrated speckle-noise-free complex amplitude imaging for both static and moving objects with tailored magnification ratio and field-of-view. The miniaturized and robust system may open the door for complex amplitude imaging in portable devices for point-of-care applications.

physics.optics↗
arXiv2023

Nancy Grace Roman Space Telescope Coronagraph Instrument Overview and Status

The Nancy Grace Roman Space Telescope Coronagraph Instrument is a critical technology demonstrator for NASA's Habitable Worlds Observatory. With a predicted visible-light flux ratio detection limit of 1E-8 or better, it will be capable of reaching new areas of parameter space for both gas giant exoplanets and circumstellar disks. It is in the final stages of integration and test at the Jet Propulsion Laboratory, with an anticipated delivery to payload integration in the coming year. This paper will review the instrument systems, observing modes, potential observing applications, and overall progress toward instrument integration and test.

astro-ph.IM↗
arXiv2023

Empowering Low-Light Image Enhancer through Customized Learnable Priors

Deep neural networks have achieved remarkable progress in enhancing low-light images by improving their brightness and eliminating noise. However, most existing methods construct end-to-end mapping networks heuristically, neglecting the intrinsic prior of image enhancement task and lacking transparency and interpretability. Although some unfolding solutions have been proposed to relieve these issues, they rely on proximal operator networks that deliver ambiguous and implicit priors. In this work, we propose a paradigm for low-light image enhancement that explores the potential of customized learnable priors to improve the transparency of the deep unfolding paradigm. Motivated by the powerful feature representation capability of Masked Autoencoder (MAE), we customize MAE-based illumination and noise priors and redevelop them from two perspectives: 1) \textbf{structure flow}: we train the MAE from a normal-light image to its illumination properties and then embed it into the proximal operator design of the unfolding architecture; and m2) \textbf{optimization flow}: we train MAE from a normal-light image to its gradient representation and then employ it as a regularization term to constrain noise in the model output. These designs improve the interpretability and representation capability of the model.Extensive experiments on multiple low-light image enhancement datasets demonstrate the superiority of our proposed paradigm over state-of-the-art methods. Code is available at https://github.com/zheng980629/CUE.

cs.CV↗
arXiv2023

FrozenRecon: Pose-free 3D Scene Reconstruction with Frozen Depth Models

3D scene reconstruction is a long-standing vision task. Existing approaches can be categorized into geometry-based and learning-based methods. The former leverages multi-view geometry but can face catastrophic failures due to the reliance on accurate pixel correspondence across views. The latter was proffered to mitigate these issues by learning 2D or 3D representation directly. However, without a large-scale video or 3D training data, it can hardly generalize to diverse real-world scenarios due to the presence of tens of millions or even billions of optimization parameters in the deep network. Recently, robust monocular depth estimation models trained with large-scale datasets have been proven to possess weak 3D geometry prior, but they are insufficient for reconstruction due to the unknown camera parameters, the affine-invariant property, and inter-frame inconsistency. Here, we propose a novel test-time optimization approach that can transfer the robustness of affine-invariant depth models such as LeReS to challenging diverse scenes while ensuring inter-frame consistency, with only dozens of parameters to optimize per video frame. Specifically, our approach involves freezing the pre-trained affine-invariant depth model's depth predictions, rectifying them by optimizing the unknown scale-shift values with a geometric consistency alignment module, and employing the resulting scale-consistent depth maps to robustly obtain camera poses and achieve dense scene reconstruction, even in low-texture regions. Experiments show that our method achieves state-of-the-art cross-dataset reconstruction on five zero-shot testing datasets.

cs.CV↗
arXiv2023

Guided Patch-Grouping Wavelet Transformer with Spatial Congruence for Ultra-High Resolution Segmentation

Most existing ultra-high resolution (UHR) segmentation methods always struggle in the dilemma of balancing memory cost and local characterization accuracy, which are both taken into account in our proposed Guided Patch-Grouping Wavelet Transformer (GPWFormer) that achieves impressive performances. In this work, GPWFormer is a Transformer ($\mathcal{T}$)-CNN ($\mathcal{C}$) mutual leaning framework, where $\mathcal{T}$ takes the whole UHR image as input and harvests both local details and fine-grained long-range contextual dependencies, while $\mathcal{C}$ takes downsampled image as input for learning the category-wise deep context. For the sake of high inference speed and low computation complexity, $\mathcal{T}$ partitions the original UHR image into patches and groups them dynamically, then learns the low-level local details with the lightweight multi-head Wavelet Transformer (WFormer) network. Meanwhile, the fine-grained long-range contextual dependencies are also captured during this process, since patches that are far away in the spatial domain can also be assigned to the same group. In addition, masks produced by $\mathcal{C}$ are utilized to guide the patch grouping process, providing a heuristics decision. Moreover, the congruence constraints between the two branches are also exploited to maintain the spatial consistency among the patches. Overall, we stack the multi-stage process in a pyramid way. Experiments show that GPWFormer outperforms the existing methods with significant improvements on five benchmark datasets.

cs.CV↗
arXiv2023

Coverage Enhancement Strategy in WMSNs Based on a Novel Swarm Intelligence Algorithm: Army Ant Search Optimizer

As one of the most crucial scenarios of the Internet of Things (IoT), wireless multimedia sensor networks (WMSNs) pay more attention to the information-intensive data (e.g., audio, video, image) for remote environments. The area coverage reflects the perception of WMSNs to the surrounding environment, where a good coverage effect can ensure effective data collection. Given the harsh and complex physical environment of WMSNs, which easily form the sensing overlapping regions and coverage holes by random deployment. The intention of our research is to deal with the optimization problem of maximizing the coverage rate in WMSNs. By proving the NP-hard of the coverage enhancement of WMSNs, inspired by the predation behavior of army ants, this article proposes a novel swarm intelligence (SI) technology army ant search optimizer (AASO) to solve the above problem, which is implemented by five operators: army ant and prey initialization, recruited by prey, attack prey, update prey, and build ant bridge. The simulation results demonstrate that the optimizer shows good performance in terms of exploration and exploitation on benchmark suites when compared to other representative SI algorithms. More importantly, coverage enhancement AASO-based in WMSNs has better merits in terms of coverage effect when compared to existing approaches.

eess.SP↗
arXiv2023

Cooperative IoT Data Sharing with Heterogeneity of Participants Based on Electricity Retail

With the development of Internet of Things (IoT) and big data technology, the data value is increasingly explored in multiple practical scenarios, including electricity transactions. However, the isolation of IoT data among several entities makes it difficult to achieve optimal allocation of data resources and convert data resources into real economic value, thus it is necessary to introduce the IoT data sharing mode to drive data circulation. To enhance the accuracy and fairness of IoT data sharing, the heterogeneity of participants is sufficiently considered, and data valuation and profit allocation in IoT data sharing are improved based on the background of electricity retail. Data valuation is supposed to be relevant to attributes of IoT data buyers, thus risk preferences of electricity retailers are applied as characteristic attributes and data premium rates are proposed to modify data value rates. Profit allocation should measure the marginal contribution shares of electricity retailers and data brokers fairly, thus asymmetric Nash bargaining model is used to guarantee that they could receive reasonable profits based on their specific contribution to the coalition of IoT data sharing. Considering the heterogeneity of participants comprehensively, the proposed IoT data sharing fits for a large coalition of IoT data sharing with multiple electricity retailers and data brokers. Finally, to demonstrate the applications of IoT data sharing in smart grids, case studies are utilized to validate the results of data value for electricity retailers with different risk preferences and the efficiency of profit allocation using asymmetric Nash bargaining model.

eess.SY↗
arXiv2023

Ultra-High Resolution Segmentation with Ultra-Rich Context: A Novel Benchmark

With the increasing interest and rapid development of methods for Ultra-High Resolution (UHR) segmentation, a large-scale benchmark covering a wide range of scenes with full fine-grained dense annotations is urgently needed to facilitate the field. To this end, the URUR dataset is introduced, in the meaning of Ultra-High Resolution dataset with Ultra-Rich Context. As the name suggests, URUR contains amounts of images with high enough resolution (3,008 images of size 5,120x5,120), a wide range of complex scenes (from 63 cities), rich-enough context (1 million instances with 8 categories) and fine-grained annotations (about 80 billion manually annotated pixels), which is far superior to all the existing UHR datasets including DeepGlobe, Inria Aerial, UDD, etc.. Moreover, we also propose WSDNet, a more efficient and effective framework for UHR segmentation especially with ultra-rich context. Specifically, multi-level Discrete Wavelet Transform (DWT) is naturally integrated to release computation burden while preserve more spatial details, along with a Wavelet Smooth Loss (WSL) to reconstruct original structured context and texture with a smooth constrain. Experiments on several UHR datasets demonstrate its state-of-the-art performance. The dataset is available at https://github.com/jankyee/URUR.

cs.CV↗
arXiv2023

MIS: A Multi-Identifier Management and Resolution System Based on Consortium Blockchain in Metaverse

The metaverse gradually evolves into a virtual world containing a series of interconnected sub-metaverses. Diverse digital resources, including identities, contents, services, and supporting data, are key components of the sub-metaverse. Therefore, a Domain Name System (DNS)-like system is necessary for efficient management and resolution. However, the legacy DNS was designed with security vulnerabilities and trust risks due to centralized issues. Blockchain is used to mitigate these concerns due to its decentralized features. Additionally, it supports identity management as a default feature, making it a natural fit for the metaverse. While there are several DNS alternatives based on the blockchain, they either manage only a single type of identifiers or isolate identities from other sorts of identifiers, making it difficult for sub-metaverses to coexist and connect with each other. This paper proposes a Multi-Identifier management and resolution System (MIS) in the metaverse, supporting the registration, resolution, and inter-translation functions. The basic MIS is portrayed as a four-tier architecture on a consortium blockchain due to its manageability, enhanced security, and efficiency properties. On-chain data is lightweight and compressed to save on storage while accelerating reading and writing operations. The resource data is encrypted based on the attributes of the sub-metaverse in the storage tier for privacy protection and access control. For users with decentralization priorities, a modification named EMIS is built on top of Ethereum. Finally, MIS is implemented on two testbeds and is available online as the open-source system. The first testbed consists of 4 physical servers located in the UK and Malaysia while the second is made up of 200 virtual machines (VMs) spread over 26 countries across all 5 continents on Google Cloud.

cs.NI↗