Large language models (LLMs) have been applied across various domains due to their superior natural language processing and generation capabilities. Nonetheless, LLMs occasionally generate content that contradicts real-world facts, known as hallucinations, posing significant challenges for real-world applications. To enhance the reliability of LLMs, it is imperative to detect hallucinations within LLM generations. Approaches that retrieve external knowledge or inspect the internal states of the model are frequently used to detect hallucinations; however, this requires either white-box access to the LLM or reliable expert knowledge resources, raising a high barrier for end-users. To address these challenges, we propose a black-box zero-resource approach for detecting LLM hallucinations, which primarily leverages multi-perspective consistency checking. The proposed approach mitigates the LLM overconfidence phenomenon by integrating multi-perspective consistency scores from both queries and responses. In comparison to the single-perspective detection approach, our proposed approach demonstrates superior performance in detecting hallucinations across multiple datasets and LLMs. Notably, in one experiment, where the hallucination rate reaches 94.7%, our approach improves the balanced accuracy (B-ACC) by 2.3 percentage points compared with the single consistency approach and achieves an area under the curve (AUC) of 0.832, all without depending on any external resources.

Optimizing video inference efficiency has become increasingly important with the growing demand for video analysis in various fields. Some existing methods achieve high efficiency by explicit discard of spatial or temporal information, which poses challenges in fast-changing and fine-grained scenarios. To address these issues, we propose an efficient video representation network with Differentiable Resolution Compression and Alignment, which compresses non-essential information in the early stage of the network to reduce computational costs while maintaining consistent temporal correlations. Specifically, we leverage a Differentiable Context-aware Compression Module to encode the saliency and non-saliency frame features, refining and updating the features into a high-low resolution video sequence. To process the new sequence, we introduce a new Resolution-Align Transformer Layer to capture global temporal correlations among frame features with different resolutions, while reducing spatial computation costs quadratically by utilizing fewer spatial tokens in low-resolution non-saliency frames. The entire network can be end-to-end optimized via the integration of the differentiable compression module. Experimental results show that our method achieves the best trade-off between efficiency and performance on near-duplicate video retrieval and competitive results on dynamic video classification compared to state-of-the-art methods.Code:https://github.com/dun-research/DRCA

Recently, to mitigate the confusion between different languages in code-switching (CS) automatic speech recognition (ASR), the conditionally factorized models, such as the language-aware encoder (LAE), explicitly disregard the contextual information between different languages. However, this information may be helpful for ASR modeling. To alleviate this issue, we propose the LAE-ST-MoE framework. It incorporates speech translation (ST) tasks into LAE and utilizes ST to learn the contextual information between different languages. It introduces a task-based mixture of expert modules, employing separate feed-forward networks for the ASR and ST tasks. Experimental results on the ASRU 2019 Mandarin-English CS challenge dataset demonstrate that, compared to the LAE-based CTC, the LAE-ST-MoE model achieves a 9.26 % mix error reduction on the CS test with the same decoding parameter. Moreover, the well-trained LAE-ST-MoE model can perform ST tasks from CS speech to Mandarin or English text.

With the development of robotics, ground robots are no longer limited to planar motion. Passive height variation due to complex terrain and active height control provided by special structures on robots require a more general navigation planning framework beyond 2D. Existing methods rarely considers both simultaneously, limiting the capabilities and applications of ground robots. In this paper, we proposed an optimization-based planning framework for ground robots considering both active and passive height changes on the z-axis. The proposed planner first constructs a penalty field for chassis motion constraints defined in$\mathbb{R}^{3}$such that the optimal solution space of the trajectory is continuous, resulting in a high-quality smooth chassis trajectory. Also, by constructing custom constraints in the z-axis direction, it is possible to plan trajectories for different types of ground robots which have z-axis degree of freedom. We performed simulations and real-world experiments to verify the efficiency and trajectory quality of our algorithm.

In recent years, advanced information communication technology (such as beyond 5G and 6G) will enable convenient end-to-end communication. This would allow collections of individual preferences as predictions of travellers’ perceptions, and thus support data analysis and traffic control in a dynamic circumstance. Therefore, analysis of nonlinearly modelled perceptions and perception-based user equilibrium are the necessary problem we encounter. This research explores the existence characteristics of perception-based stochastic user equilibrium (P-SUE), a type of stochastic user equilibrium where travelers make route choices according to their perceptive traffic circumstances instead of actual ones. A nonlinear perception model with stochasticity is developed and the traffic arrivals are approximately transformed from Poisson-distributed to Normal-distributed. From a static perspective, we develop a P-SUE model with a heterogeneous traveler community and discuss the interaction among travelers’ attitude, grouping and static perception-based stochastic user equilibrium (P-SSUE). After embedding a day-to-day learning process into the P-SSUE model, we further explore evolutionary dynamics of the equilibrium. Several scenarios are provided to investigate effects and sensitivity of traveler-related and road-related factors. It is found that learning produces perturbation and makes the existence condition of P-SSUE more complicated; the convergence processes of P-SSUEs are significantly shortened as the learning rate increases to higher than 0.1.