arXiv subjects

Yi Yang

Publications and source records attributed to Yi Yang.

At least 577 records · Page 32Linked to original sources

arXiv2024

LSK3DNet: Towards Effective and Efficient 3D Perception with Large Sparse Kernels

Autonomous systems need to process large-scale, sparse, and irregular point clouds with limited compute resources. Consequently, it is essential to develop LiDAR perception methods that are both efficient and effective. Although naively enlarging 3D kernel size can enhance performance, it will also lead to a cubically-increasing overhead. Therefore, it is crucial to develop streamlined 3D large kernel designs that eliminate redundant weights and work effectively with larger kernels. In this paper, we propose an efficient and effective Large Sparse Kernel 3D Neural Network (LSK3DNet) that leverages dynamic pruning to amplify the 3D kernel size. Our method comprises two core components: Spatial-wise Dynamic Sparsity (SDS) and Channel-wise Weight Selection (CWS). SDS dynamically prunes and regrows volumetric weights from the beginning to learn a large sparse 3D kernel. It not only boosts performance but also significantly reduces model size and computational cost. Moreover, CWS selects the most important channels for 3D convolution during training and subsequently prunes the redundant channels to accelerate inference for 3D vision tasks. We demonstrate the effectiveness of LSK3DNet on three benchmark datasets and five tracks compared with classical models and large kernel designs. Notably, LSK3DNet achieves the state-of-the-art performance on SemanticKITTI (i.e., 75.6% on single-scan and 63.4% on multi-scan), with roughly 40% model size reduction and 60% computing operations reduction compared to the naive large 3D kernel model.

cs.CV↗

arXiv2024

Volumetric Environment Representation for Vision-Language Navigation

Vision-language navigation (VLN) requires an agent to navigate through an 3D environment based on visual observations and natural language instructions. It is clear that the pivotal factor for successful navigation lies in the comprehensive scene understanding. Previous VLN agents employ monocular frameworks to extract 2D features of perspective views directly. Though straightforward, they struggle for capturing 3D geometry and semantics, leading to a partial and incomplete environment representation. To achieve a comprehensive 3D representation with fine-grained details, we introduce a Volumetric Environment Representation (VER), which voxelizes the physical world into structured 3D cells. For each cell, VER aggregates multi-view 2D features into such a unified 3D space via 2D-3D sampling. Through coarse-to-fine feature extraction and multi-task learning for VER, our agent predicts 3D occupancy, 3D room layout, and 3D bounding boxes jointly. Based on online collected VERs, our agent performs volume state estimation and builds episodic memory for predicting the next step. Experimental results show our environment representations from multi-task learning lead to evident performance gains on VLN. Our model achieves state-of-the-art performance across VLN benchmarks (R2R, REVERIE, and R4R).

cs.CV↗

arXiv2024

Beyond Surface Similarity: Detecting Subtle Semantic Shifts in Financial Narratives

In this paper, we introduce the Financial-STS task, a financial domain-specific NLP task designed to measure the nuanced semantic similarity between pairs of financial narratives. These narratives originate from the financial statements of the same company but correspond to different periods, such as year-over-year comparisons. Measuring the subtle semantic differences between these paired narratives enables market stakeholders to gauge changes over time in the company's financial and operational situations, which is critical for financial decision-making. We find that existing pretrained embedding models and LLM embeddings fall short in discerning these subtle financial narrative shifts. To address this gap, we propose an LLM-augmented pipeline specifically designed for the Financial-STS task. Evaluation on a human-annotated dataset demonstrates that our proposed method outperforms existing methods trained on classic STS tasks and generic LLM embeddings.

cs.CL↗

arXiv2024

Knowledge-guided Causal Intervention for Weakly-supervised Object Localization

Previous weakly-supervised object localization (WSOL) methods aim to expand activation map discriminative areas to cover the whole objects, yet neglect two inherent challenges when relying solely on image-level labels. First, the ``entangled context'' issue arises from object-context co-occurrence (\eg, fish and water), making the model inspection hard to distinguish object boundaries clearly. Second, the ``C-L dilemma'' issue results from the information decay caused by the pooling layers, which struggle to retain both the semantic information for precise classification and those essential details for accurate localization, leading to a trade-off in performance. In this paper, we propose a knowledge-guided causal intervention method, dubbed KG-CI-CAM, to address these two under-explored issues in one go. More specifically, we tackle the co-occurrence context confounder problem via causal intervention, which explores the causalities among image features, contexts, and categories to eliminate the biased object-context entanglement in the class activation maps. Based on the disentangled object feature, we introduce a multi-source knowledge guidance framework to strike a balance between absorbing classification knowledge and localization knowledge during model training. Extensive experiments conducted on several benchmark datasets demonstrate the effectiveness of KG-CI-CAM in learning distinct object boundaries amidst confounding contexts and mitigating the dilemma between classification and localization performance.

cs.CV↗

arXiv2024

Entangled View-Epipolar Information Aggregation for Generalizable Neural Radiance Fields

Generalizable NeRF can directly synthesize novel views across new scenes, eliminating the need for scene-specific retraining in vanilla NeRF. A critical enabling factor in these approaches is the extraction of a generalizable 3D representation by aggregating source-view features. In this paper, we propose an Entangled View-Epipolar Information Aggregation method dubbed EVE-NeRF. Different from existing methods that consider cross-view and along-epipolar information independently, EVE-NeRF conducts the view-epipolar feature aggregation in an entangled manner by injecting the scene-invariant appearance continuity and geometry consistency priors to the aggregation process. Our approach effectively mitigates the potential lack of inherent geometric and appearance constraint resulting from one-dimensional interactions, thus further boosting the 3D representation generalizablity. EVE-NeRF attains state-of-the-art performance across various evaluation scenarios. Extensive experiments demonstate that, compared to prevailing single-dimensional aggregation, the entangled network excels in the accuracy of 3D scene geometry and appearance reconstruction. Our code is publicly available at https://github.com/tatakai1/EVENeRF.

cs.CV↗

arXiv2024

Counterfactual Co-occurring Learning for Bias Mitigation in Weakly-supervised Object Localization

Contemporary weakly-supervised object localization (WSOL) methods have primarily focused on addressing the challenge of localizing the most discriminative region while largely overlooking the relatively less explored issue of biased activation -- incorrectly spotlighting co-occurring background with the foreground feature. In this paper, we conduct a thorough causal analysis to investigate the origins of biased activation. Based on our analysis, we attribute this phenomenon to the presence of co-occurring background confounders. Building upon this profound insight, we introduce a pioneering paradigm known as Counterfactual Co-occurring Learning (CCL), meticulously engendering counterfactual representations by adeptly disentangling the foreground from the co-occurring background elements. Furthermore, we propose an innovative network architecture known as Counterfactual-CAM. This architecture seamlessly incorporates a perturbation mechanism for counterfactual representations into the vanilla CAM-based model. By training the WSOL model with these perturbed representations, we guide the model to prioritize the consistent foreground content while concurrently reducing the influence of distracting co-occurring backgrounds. To the best of our knowledge, this study represents the initial exploration of this research direction. Our extensive experiments conducted across multiple benchmarks validate the effectiveness of the proposed Counterfactual-CAM in mitigating biased activation.

cs.CV↗

arXiv2024

DLP-GAN: learning to draw modern Chinese landscape photos with generative adversarial network

Chinese landscape painting has a unique and artistic style, and its drawing technique is highly abstract in both the use of color and the realistic representation of objects. Previous methods focus on transferring from modern photos to ancient ink paintings. However, little attention has been paid to translating landscape paintings into modern photos. To solve such problems, in this paper, we (1) propose DLP-GAN (Draw Modern Chinese Landscape Photos with Generative Adversarial Network), an unsupervised cross-domain image translation framework with a novel asymmetric cycle mapping, and (2) introduce a generator based on a dense-fusion module to match different translation directions. Moreover, a dual-consistency loss is proposed to balance the realism and abstraction of model painting. In this way, our model can draw landscape photos and sketches in the modern sense. Finally, based on our collection of modern landscape and sketch datasets, we compare the images generated by our model with other benchmarks. Extensive experiments including user studies show that our model outperforms state-of-the-art methods.

cs.CV↗

arXiv2024

AntEval: Evaluation of Social Interaction Competencies in LLM-Driven Agents

Large Language Models (LLMs) have demonstrated their ability to replicate human behaviors across a wide range of scenarios. However, their capability in handling complex, multi-character social interactions has yet to be fully explored, primarily due to the absence of robust, quantitative evaluation methods. This gap has slowed the development of agents proficient in more nuanced interactions beyond simple exchanges, for example, small talk. To address this challenge, we introduce the Multi-Agent Interaction Evaluation Framework (AntEval), encompassing a novel interaction framework and evaluation methods. The interaction framework aims to foster an complex interaction environment that bolsters information exchange and intention expression within social interactions. Furthermore, we introduce evaluation methods, including two metrics: Information Exchanging Precision (IEP) and Interaction Expressiveness Gap (IEG), designed for the quantitative and objective assessment of agents' interaction competencies. Our findings highlight the utility of these evaluative methods and show significant potential for improving LLMs' ability to construct agents that interact in a more natural manner with human-like intricacy.

cs.CL↗

arXiv2024

MIGC: Multi-Instance Generation Controller for Text-to-Image Synthesis

We present a Multi-Instance Generation (MIG) task, simultaneously generating multiple instances with diverse controls in one image. Given a set of predefined coordinates and their corresponding descriptions, the task is to ensure that generated instances are accurately at the designated locations and that all instances' attributes adhere to their corresponding description. This broadens the scope of current research on Single-instance generation, elevating it to a more versatile and practical dimension. Inspired by the idea of divide and conquer, we introduce an innovative approach named Multi-Instance Generation Controller (MIGC) to address the challenges of the MIG task. Initially, we break down the MIG task into several subtasks, each involving the shading of a single instance. To ensure precise shading for each instance, we introduce an instance enhancement attention mechanism. Lastly, we aggregate all the shaded instances to provide the necessary information for accurately generating multiple instances in stable diffusion (SD). To evaluate how well generation models perform on the MIG task, we provide a COCO-MIG benchmark along with an evaluation pipeline. Extensive experiments were conducted on the proposed COCO-MIG benchmark, as well as on various commonly used benchmarks. The evaluation results illustrate the exceptional control capabilities of our model in terms of quantity, position, attribute, and interaction. Code and demos will be released at https://migcproject.github.io/.

cs.CV↗

arXiv2024

Multiple-environment Self-adaptive Network for Aerial-view Geo-localization

Aerial-view geo-localization tends to determine an unknown position through matching the drone-view image with the geo-tagged satellite-view image. This task is mostly regarded as an image retrieval problem. The key underpinning this task is to design a series of deep neural networks to learn discriminative image descriptors. However, existing methods meet large performance drops under realistic weather, such as rain and fog, since they do not take the domain shift between the training data and multiple test environments into consideration. To minor this domain gap, we propose a Multiple-environment Self-adaptive Network (MuSe-Net) to dynamically adjust the domain shift caused by environmental changing. In particular, MuSe-Net employs a two-branch neural network containing one multiple-environment style extraction network and one self-adaptive feature extraction network. As the name implies, the multiple-environment style extraction network is to extract the environment-related style information, while the self-adaptive feature extraction network utilizes an adaptive modulation module to dynamically minimize the environment-related style gap. Extensive experiments on two widely-used benchmarks, i.e., University-1652 and CVUSA, demonstrate that the proposed MuSe-Net achieves a competitive result for geo-localization in multiple environments. Furthermore, we observe that the proposed method also shows great potential to the unseen extreme weather, such as mixing the fog, rain and snow.

cs.CV↗

arXiv2024

Test-Time Adaptation with CLIP Reward for Zero-Shot Generalization in Vision-Language Models

One fascinating aspect of pre-trained vision-language models~(VLMs) learning under language supervision is their impressive zero-shot generalization capability. However, this ability is hindered by distribution shifts between the training and testing data. Previous test time adaptation~(TTA) methods for VLMs in zero-shot classification rely on minimizing the entropy of model outputs, tending to be stuck in incorrect model predictions. In this work, we propose TTA with feedback to rectify the model output and prevent the model from becoming blindly confident. Specifically, a CLIP model is adopted as the reward model during TTA and provides feedback for the VLM. Given a single test sample, the VLM is forced to maximize the CLIP reward between the input and sampled results from the VLM output distribution. The proposed \textit{reinforcement learning with CLIP feedback~(RLCF)} framework is highly flexible and universal. Beyond the classification task, with task-specific sampling strategies and a proper reward baseline choice, RLCF can be easily extended to not only discrimination tasks like retrieval but also generalization tasks like image captioning, improving the zero-shot generalization capacity of VLMs. According to the characteristics of these VL tasks, we build different fully TTA pipelines with RLCF to improve the zero-shot generalization ability of various VLMs. Extensive experiments along with promising empirical results demonstrate the effectiveness of RLCF. The code is available at https://github.com/mzhaoshuai/RLCF.

cs.CV↗

arXiv2024

Disentangling Structured Components: Towards Adaptive, Interpretable and Scalable Time Series Forecasting

Multivariate time-series (MTS) forecasting is a paramount and fundamental problem in many real-world applications. The core issue in MTS forecasting is how to effectively model complex spatial-temporal patterns. In this paper, we develop a adaptive, interpretable and scalable forecasting framework, which seeks to individually model each component of the spatial-temporal patterns. We name this framework SCNN, as an acronym of Structured Component-based Neural Network. SCNN works with a pre-defined generative process of MTS, which arithmetically characterizes the latent structure of the spatial-temporal patterns. In line with its reverse process, SCNN decouples MTS data into structured and heterogeneous components and then respectively extrapolates the evolution of these components, the dynamics of which are more traceable and predictable than the original MTS. Extensive experiments are conducted to demonstrate that SCNN can achieve superior performance over state-of-the-art models on three real-world datasets. Additionally, we examine SCNN with different configurations and perform in-depth analyses of the properties of SCNN.

cs.LG↗

arXiv2024

Do LLMs Know about Hallucination? An Empirical Investigation of LLM's Hidden States

Large Language Models (LLMs) can make up answers that are not real, and this is known as hallucination. This research aims to see if, how, and to what extent LLMs are aware of hallucination. More specifically, we check whether and how an LLM reacts differently in its hidden states when it answers a question right versus when it hallucinates. To do this, we introduce an experimental framework which allows examining LLM's hidden states in different hallucination situations. Building upon this framework, we conduct a series of experiments with language models in the LLaMA family (Touvron et al., 2023). Our empirical findings suggest that LLMs react differently when processing a genuine response versus a fabricated one. We then apply various model interpretation techniques to help understand and explain the findings better. Moreover, informed by the empirical observations, we show great potential of using the guidance derived from LLM's hidden representation space to mitigate hallucination. We believe this work provides insights into how LLMs produce hallucinated answers and how to make them occur less often.

cs.CL↗

arXiv2024

A seven-Earth-radius helium-burning star inside a 20.5-min detached binary

Binary evolution theory predicts that the second common envelope (CE) ejection can produce low-mass (0.32-0.36 Msun) subdwarf B (sdB) stars inside ultrashort-orbital-period binary systems, as their helium cores are ignited under nondegenerate conditions. With the orbital decay driven by gravitational-wave (GW) radiation, the minimum orbital periods of detached sdB binaries could be as short as ~20 minutes. However, only four sdB binaries with orbital periods below an hour have been reported so far, while none of them has an orbital period approaching the above theoretical limit. Here we report the discovery of a 20.5-minute-orbital-period ellipsoidal binary, TMTS J052610.43+593445.1, in which the visible star is being tidally deformed by an invisible carbon-oxygen white dwarf (WD) companion. The visible component is inferred to be an sdB star with a mass of ~0.33 Msun, approaching that of helium-ignition limit, although a He-core WD cannot be completely ruled out. In particular, the radius of this low-mass sdB star is only 0.066 Rsun, about seven Earth radii, possibly representing the most compact nondegenerate star ever known. Such a system provides a key clue to map the binary evolution scheme from the second CE ejection to the formation of AM CVn stars having a helium-star donor, and it will also serve as a crucial verification binary of space-borne GW detectors in the future.

astro-ph.SR↗

arXiv2024

Modeling the black holes surrounded by a dark matter halo in the galactic center of M87

In this paper, the structure of a dark matter halo can be well described by the mass model of M87 and the Einasto profile for the cold dark matter model, i.e., $ρ_{\text{eina}} (r)=ρ_\text{e} \exp ( -2 α^{-1} ((r/r_\text{e})^α-1 ) )$ (Wang et al. in Nature 585:39-42, 2020). Under these conditions, we construct a solution of a static spherically symmetric black hole in a dark matter halo. Then, using the Newman-janis algorithm, we extend this static solution to the case of rotation, and obtain a solution for the Kerr-like black hole. We prove that this solution of the Kerr-like black hole is indeed a solution to the Einstein field equations. Finally, taking M87 as an example, we study and analyze some physical properties of this Kerr-like black hole, and then compare them with the Kerr black hole. Particularly, from the perspective of the black hole shadow and the fact that the Kerr-like black hole and the Kerr black hole is distinguishable, we give the upper limit of the shape parameter of the Einasto density profile, that is approximately $α<0.22$, which may provide a new method to further improve and perfect the density profile of dark matter model. These research results for the black hole in a dark matter halo may indirectly provide an effective method for detecting the existence of dark matter.

gr-qc↗

arXiv2024

Retrosynthesis prediction enhanced by in-silico reaction data augmentation

Recent advances in machine learning (ML) have expedited retrosynthesis research by assisting chemists to design experiments more efficiently. However, all ML-based methods consume substantial amounts of paired training data (i.e., chemical reaction: product-reactant(s) pair), which is costly to obtain. Moreover, companies view reaction data as a valuable asset and restrict the accessibility to researchers. These issues prevent the creation of more powerful retrosynthesis models due to their data-driven nature. As a response, we exploit easy-to-access unpaired data (i.e., one component of product-reactant(s) pair) for generating in-silico paired data to facilitate model training. Specifically, we present RetroWISE, a self-boosting framework that employs a base model inferred from real paired data to perform in-silico reaction generation and augmentation using unpaired data, ultimately leading to a superior model. On three benchmark datasets, RetroWISE achieves the best overall performance against state-of-the-art models (e.g., +8.6% top-1 accuracy on the USPTO-50K test dataset). Moreover, it consistently improves the prediction accuracy of rare transformations. These results show that Retro- WISE overcomes the training bottleneck by in-silico reactions, thereby paving the way toward more effective ML-based retrosynthesis models.

cs.LG↗

arXiv2024

DeFlow: Decoder of Scene Flow Network in Autonomous Driving

Scene flow estimation determines a scene's 3D motion field, by predicting the motion of points in the scene, especially for aiding tasks in autonomous driving. Many networks with large-scale point clouds as input use voxelization to create a pseudo-image for real-time running. However, the voxelization process often results in the loss of point-specific features. This gives rise to a challenge in recovering those features for scene flow tasks. Our paper introduces DeFlow which enables a transition from voxel-based features to point features using Gated Recurrent Unit (GRU) refinement. To further enhance scene flow estimation performance, we formulate a novel loss function that accounts for the data imbalance between static and dynamic points. Evaluations on the Argoverse 2 scene flow task reveal that DeFlow achieves state-of-the-art results on large-scale point cloud data, demonstrating that our network has better performance and efficiency compared to others. The code is open-sourced at https://github.com/KTH-RPL/deflow.

cs.CV↗

arXiv2024

MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries

Retrieval-augmented generation (RAG) augments large language models (LLM) by retrieving relevant knowledge, showing promising potential in mitigating LLM hallucinations and enhancing response quality, thereby facilitating the great adoption of LLMs in practice. However, we find that existing RAG systems are inadequate in answering multi-hop queries, which require retrieving and reasoning over multiple pieces of supporting evidence. Furthermore, to our knowledge, no existing RAG benchmarking dataset focuses on multi-hop queries. In this paper, we develop a novel dataset, MultiHop-RAG, which consists of a knowledge base, a large collection of multi-hop queries, their ground-truth answers, and the associated supporting evidence. We detail the procedure of building the dataset, utilizing an English news article dataset as the underlying RAG knowledge base. We demonstrate the benchmarking utility of MultiHop-RAG in two experiments. The first experiment compares different embedding models for retrieving evidence for multi-hop queries. In the second experiment, we examine the capabilities of various state-of-the-art LLMs, including GPT-4, PaLM, and Llama2-70B, in reasoning and answering multi-hop queries given the evidence. Both experiments reveal that existing RAG methods perform unsatisfactorily in retrieving and answering multi-hop queries. We hope MultiHop-RAG will be a valuable resource for the community in developing effective RAG systems, thereby facilitating greater adoption of LLMs in practice. The MultiHop-RAG and implemented RAG system is publicly available at https://github.com/yixuantt/MultiHop-RAG/.

cs.CL↗