arXiv subjects
Publications and source records attributed to Li He.
After the universal property of the six functor formalism $\Shv(-;\Sp)$ on locally compact Hausdorff spaces given by Zhu, we show that the six functor formalism $\Shv(-;\Sp)$ on light condensed anima in the sense of Heyer-Mann is initial among all six functor formalisms $D$ satisfying some mild conditions, and then we present some applications.
Obstacle avoidance in complex and dynamic environments is a critical challenge for real-time robot navigation. Model-based and learning-based methods often fail in highly dynamic scenarios because traditional methods assume a static environment and cannot adapt to real-time changes, while learning-based methods rely on single-frame observations for motion constraint estimation, limiting their adaptability. To overcome these limitations, this paper proposes a novel framework that leverages multi-frame point constraints, including current and future frames predicted by a dedicated module, to enable proactive end-to-end navigation. By incorporating a prediction module that forecasts the future path of moving obstacles based on multi-frame observations, our method allows the robot to proactively anticipate and avoid potential dangers. This proactive planning capability significantly enhances navigation robustness and efficiency in unknown dynamic environments. Simulations and real-world experiments validate the effectiveness of our approach.
We give an internal description of constructible objects in an $\infty$-topos. More precisely, $P$-consctructible objects are locally constant objects internal to Fun($P$,An), for any noetherian poset $P$.
Recent advances in correlation-based sequential recommendation systems have demonstrated substantial success. Specifically, the attention-based model outperforms other RNN-based and Markov chains-based models by capturing both short- and long-term dependencies more effectively. However, solely focusing on item co-occurrences overlooks the underlying motivations behind user behaviors, leading to spurious correlations and potentially inaccurate recommendations. To address this limitation, we present a novel framework that integrates causal attention for sequential recommendation, CausalRec. It incorporates a causal discovery block and a CausalBooster. The causal discovery block learns the causal graph in user behavior sequences, and we provide a theory to guarantee the identifiability of the learned causal graph. The CausalBooster utilizes the discovered causal graph to refine the attention mechanism, prioritizing behaviors with causal significance. Experimental evaluations on real-world datasets indicate that CausalRec outperforms several state-of-the-art methods, with average improvements of 7.21% in Hit Rate (HR) and 8.65% in Normalized Discounted Cumulative Gain (NDCG). To the best of our knowledge, this is the first model to incorporate causality through the attention mechanism in sequential recommendation, demonstrating the value of causality in generating more accurate and reliable recommendations.
Recent visual place recognition (VPR) approaches have leveraged foundation models (FM) and introduced novel aggregation techniques. However, these methods have failed to fully exploit key concepts of FM, such as the effective utilization of extensive training sets, and they have overlooked the potential of classical aggregation methods, such as GeM and NetVLAD. Building on these insights, we revive classical feature aggregation methods and develop more fundamental VPR models, collectively termed SuperPlace. First, we introduce a supervised label alignment method that enables training across various VPR datasets within a unified framework. Second, we propose G$^2$M, a compact feature aggregation method utilizing two GeMs, where one GeM learns the principal components of feature maps along the channel dimension and calibrates the output of the other. Third, we propose the secondary fine-tuning (FT$^2$) strategy for NetVLAD-Linear (NVL). NetVLAD first learns feature vectors in a high-dimensional space and then compresses them into a lower-dimensional space via a single linear layer. Extensive experiments highlight our contributions and demonstrate the superiority of SuperPlace. Specifically, G$^2$M achieves promising results with only one-tenth of the feature dimensions compared to recent methods. Moreover, NVL-FT$^2$ ranks first on the MSLS leaderboard.
Online AI Feedback (OAIF) presents a promising alternative to Reinforcement Learning from Human Feedback (RLHF) by utilizing online AI preference in aligning language models (LLMs). However, the straightforward replacement of humans with AI deprives LLMs from learning more fine-grained AI supervision beyond binary signals. In this paper, we propose Direct Advantage Regression (DAR), a simple alignment algorithm using online AI reward to optimize policy improvement through weighted supervised fine-tuning. As an RL-free approach, DAR maintains theoretical consistency with online RLHF pipelines while significantly reducing implementation complexity and improving learning efficiency. Our empirical results underscore that AI reward is a better form of AI supervision consistently achieving higher human-AI agreement as opposed to AI preference. Additionally, evaluations using GPT-4-Turbo and MT-bench show that DAR outperforms both OAIF and online RLHF baselines.
Long-form article generation (LFAG) presents challenges such as maintaining logical consistency, comprehensive topic coverage, and narrative coherence across extended articles. Existing datasets often lack both the hierarchical structure and fine-grained annotation needed to effectively decompose tasks, resulting in shallow, disorganized article generation. To address these limitations, we introduce DeFine, a Decomposed and Fine-grained annotated dataset for long-form article generation. DeFine is characterized by its hierarchical decomposition strategy and the integration of domain-specific knowledge with multi-level annotations, ensuring granular control and enhanced depth in article generation. To construct the dataset, a multi-agent collaborative pipeline is proposed, which systematically segments the generation process into four parts: Data Miner, Cite Retreiver, Q&A Annotator and Data Cleaner. To validate the effectiveness of DeFine, we designed and tested three LFAG baselines: the web retrieval, the local retrieval, and the grounded reference. We fine-tuned the Qwen2-7b-Instruct model using the DeFine training dataset. The experimental results showed significant improvements in text quality, specifically in topic coverage, depth of information, and content fidelity. Our dataset publicly available to facilitate future research.
With the rise of large-scale language models (LLMs), it is currently popular and effective to convert multimodal information into text descriptions for multimodal multi-hop question answering. However, we argue that the current methods of multi-modal multi-hop question answering still mainly face two challenges: 1) The retrieved evidence containing a large amount of redundant information, inevitably leads to a significant drop in performance due to irrelevant information misleading the prediction. 2) The reasoning process without interpretable reasoning steps makes the model difficult to discover the logical errors for handling complex questions. To solve these problems, we propose a unified LLMs-based approach but without heavily relying on them due to the LLM's potential errors, and innovatively treat multimodal multi-hop question answering as a joint entailment tree generation and question answering problem. Specifically, we design a multi-task learning framework with a focus on facilitating common knowledge sharing across interpretability and prediction tasks while preventing task-specific errors from interfering with each other via mixture of experts. Afterward, we design an iterative feedback mechanism to further enhance both tasks by feeding back the results of the joint training to the LLM for regenerating entailment trees, aiming to iteratively refine the potential answer. Notably, our method has won the first place in the official leaderboard of WebQA (since April 10, 2024), and achieves competitive results on MultimodalQA.
Strong coupling between light and matter in an optical cavity provides a pathway to giant polariton nonlinearity, where effective polariton-polariton interactions are mediated by materials' nonlinear responses. The pursuit of such enhanced nonlinearity at low optical excitations, potentially down to the single-particle level, has been a central focus in the field, inspiring the exploration of novel solid-state light-matter systems. Here, we experimentally realize extremely nonlinear and robust cavity exciton-polaritons by coupling a charge-tunable MoSe2 monolayer to a photonic crystal nanocavity. We show that the observed polariton nonlinearity arises from increased exciton dephasing at high populations, leading to diminished exciton-photon coupling and ultimately the breakdown of the strong coupling condition. Remarkably, the strong mode confinement of the nanocavity enables all-optical switching of the cavity spectrum at ultralow optical excitation energies, down to ~4 fJ, on picosecond timescales. Our work paves the way for further exploration of 2D nonlinear exciton-polaritons, with promising applications in both classical and quantum all-optical information processing.
Thermal emission is a ubiquitous electromagnetic wave with an extreme broad spectrum in nature, and controlling thermal emission can be used to develop low-cost and convenient infrared light sources with wavelength tunable in a wide range that is currently difficult to other sources. Conventional metasurfaces are commonly used to control light but lack the flexibility to achieve complex emission spectral profiles and dynamic tuning. Here, we introduce a novel dual-layer metasurface structure with two completely independent layers to achieve a multi-peak thermal emission within the 5-8 μm wavelength range. Simulations and experiments show that this two-layer structure can achieve arbitrary spectral shapes without interfering with multiple resonant modes. This unique configuration presents a promising platform for further exploration in thermal emission engineering, enabling spectral control and dynamic tuning.
Self-assembled molecular monolayer doping remains as a research focus for its nature of being conformal, nondestructive, and self-limiting. Herein, we demonstrate a sulfur monolayer doping in GaAs, facilitated by (NH4)2Sx solution. The Van der Pauw technique, secondary-ion mass spectroscopy, and low-temperature Hall effect measurements show that the sulfur dopants concentration and electron activation rate are 4*10^20 cm-3 and 77.6%, respectively. The donor energy level of sulfur-doped GaAs is located 68 meV below the conduction band. Based on this process, a p-n junction was successfully fabricated on highly doped p-type GaAs substrate.
Invasive brain-computer interfaces with Electrocorticography (ECoG) have shown promise for high-performance speech decoding in medical applications, but less damaging methods like intracranial stereo-electroencephalography (sEEG) remain underexplored. With rapid advances in representation learning, leveraging abundant recordings to enhance speech decoding is increasingly attractive. However, popular methods often pre-train temporal models based on brain-level tokens, overlooking that brain activities in different regions are highly desynchronized during tasks. Alternatively, they pre-train spatial-temporal models based on channel-level tokens but fail to evaluate them on challenging tasks like speech decoding, which requires intricate processing in specific language-related areas. To address this issue, we collected a well-annotated Chinese word-reading sEEG dataset targeting language-related brain networks from 12 subjects. Using this benchmark, we developed the Du-IN model, which extracts contextual embeddings based on region-level tokens through discrete codex-guided mask modeling. Our model achieves state-of-the-art performance on the 61-word classification task, surpassing all baselines. Model comparisons and ablation studies reveal that our design choices, including (i) temporal modeling based on region-level tokens by utilizing 1D depthwise convolution to fuse channels in the ventral sensorimotor cortex (vSMC) and superior temporal gyrus (STG) and (ii) self-supervision through discrete codex-guided mask modeling, significantly contribute to this performance. Overall, our approach -- inspired by neuroscience findings and capitalizing on region-level representations from specific brain regions -- is suitable for invasive brain modeling and represents a promising neuro-inspired AI approach in brain-computer interfaces.
Neural implicit surface reconstruction has achieved remarkable progress recently. Despite resorting to complex radiance modeling, state-of-the-art methods still struggle with textureless and specular surfaces. Different from RGB images, polarization images can provide direct constraints on the azimuth angles of the surface normals. In this paper, we present PISR, a novel method that utilizes a geometrically accurate polarimetric loss to refine shape independently of appearance. In addition, PISR smooths surface normals in image space to eliminate severe shape distortions and leverages the hash-grid-based neural signed distance function to accelerate the reconstruction. Experimental results demonstrate that PISR achieves higher accuracy and robustness, with an L1 Chamfer distance of 0.5 mm and an F-score of 99.5% at 1 mm, while converging 4~30 times faster than previous polarimetric surface reconstruction methods.
Robot person following (RPF) is a crucial capability in human-robot interaction (HRI) applications, allowing a robot to persistently follow a designated person. In practical RPF scenarios, the person can often be occluded by other objects or people. Consequently, it is necessary to re-identify the person when he/she reappears within the robot's field of view. Previous person re-identification (ReID) approaches to person following rely on a fixed feature extractor. Such an approach often fails to generalize to different viewpoints and lighting conditions in practical RPF environments. In other words, it suffers from the so-called domain shift problem where it cannot re-identify the person when his re-appearance is out of the domain modeled by the fixed feature extractor. To mitigate this problem, we propose a ReID framework for RPF where we use a feature extractor that is optimized online with both short-term and long-term experiences (i.e., recently and previously observed samples during RPF) using the online continual learning (OCL) framework. The long-term experiences are maintained by a memory manager to enable OCL to update the feature extractor. Our experiments demonstrate that even in the presence of severe appearance changes and distractions from visually similar people, the proposed method can still re-identify the person more accurately than the state-of-the-art methods.
Large-scale point cloud consists of a multitude of individual objects, thereby encompassing rich structural and underlying semantic contextual information, resulting in a challenging problem in efficiently segmenting a point cloud. Most existing researches mainly focus on capturing intricate local features without giving due consideration to global ones, thus failing to leverage semantic context. In this paper, we propose a Similarity-Weighted Convolution and local-global Fusion Network, named SWCF-Net, which takes into account both local and global features. We propose a Similarity-Weighted Convolution (SWConv) to effectively extract local features, where similarity weights are incorporated into the convolution operation to enhance the generalization capabilities. Then, we employ a downsampling operation on the K and V channels within the attention module, thereby reducing the quadratic complexity to linear, enabling the Transformer to deal with large-scale point clouds. At last, orthogonal components are extracted in the global features and then aggregated with local features, thereby eliminating redundant information between local and global features and consequently promoting efficiency. We evaluate SWCF-Net on large-scale outdoor datasets SemanticKITTI and Toronto3D. Our experimental results demonstrate the effectiveness of the proposed network. Our method achieves a competitive result with less computational cost, and is able to handle large-scale point clouds efficiently.
The capability of continuously learning new skills via a sequence of pre-collected offline datasets is desired for an agent. However, consecutively learning a sequence of offline tasks likely leads to the catastrophic forgetting issue under resource-limited scenarios. In this paper, we formulate a new setting, continual offline reinforcement learning (CORL), where an agent learns a sequence of offline reinforcement learning tasks and pursues good performance on all learned tasks with a small replay buffer without exploring any of the environments of all the sequential tasks. For consistently learning on all sequential tasks, an agent requires acquiring new knowledge and meanwhile preserving old knowledge in an offline manner. To this end, we introduced continual learning algorithms and experimentally found experience replay (ER) to be the most suitable algorithm for the CORL problem. However, we observe that introducing ER into CORL encounters a new distribution shift problem: the mismatch between the experiences in the replay buffer and trajectories from the learned policy. To address such an issue, we propose a new model-based experience selection (MBES) scheme to build the replay buffer, where a transition model is learned to approximate the state distribution. This model is used to bridge the distribution bias between the replay buffer and the learned model by filtering the data from offline data that most closely resembles the learned model for storage. Moreover, in order to enhance the ability on learning new tasks, we retrofit the experience replay method with a new dual behavior cloning (DBC) architecture to avoid the disturbance of behavior-cloning loss on the Q-learning process. In general, we call our algorithm offline experience replay (OER). Extensive experiments demonstrate that our OER method outperforms SOTA baselines in widely-used Mujoco environments.
Since wearable technology was first conceptualized in 1960, it has become a crucial aspect of our daily life. People's daily lives have been made better and their health and well-being have improved because of its convenience, real-time, and efficiency, among other things. Haptic feedback and haptic-based cues are crucial tools to improve interaction when trying to improve user experience. An essential component of human-computer interface design that can be used to improve user experience is haptic feedback. By offering a more immersive interface, haptic feedback can improve the user experience. This paper attempts to better understand "The role of haptic feedback in enhancing the user experience of wearable devices" and the effects of haptic feedback on user engagement and interaction. The findings of this study are relevant to the developing fields of wearable technology and human-computer interaction. Wearable technologies can be made more intuitive, user-friendly, and immersive by considering how haptic feedback affects the user experience.
Traditional static user interfaces (UI) have given way to dynamic systems that can intelligently adapt to and respond to users' changing needs. Temporal interaction is an emerging field in human-computer interaction (HCI), which refers to the study and design of UI that are capable of adapting and responding to the user's changing behavioral and emotional states. By comprehending and incorporating the temporal component of user interactions, it focuses on developing dynamic and individualized user experiences. This idea places a strong emphasis on the value of adjusting to user behavior and emotions in order to create a more unique and interesting user experience. The potential of temporal interaction to alter user interface design is highlighted by this paper's examination of its capacity to adjust to user behavior and react to emotional states. Designers can create interfaces that respond to the changing demands, emotions, and behaviors of users by utilizing temporal interactions. This produces interfaces that are not only highly functional but also form an emotional connection with the users.