Exposure bias is a notorious issue in Recommender Systems (RSs), often resulting in the overexposure of a few items while a majority of items are underexposed. Most existing methods mitigate exposure bias in offline recommendations, but few research address such bias in Interactive Recommender Systems (IRSs). However, the exposure bias will be increasingly amplified over time as users interact with the system. Worse still, bias amplification will further lead to a series of problems such as echo chambers and filter bubbles. To address these issues, we propose a novel paradigm, Causal Intervention-Inspired Policy Learning to Mitigate Exposure Bias for Interactive RECommendation (CIREC), which consists of Intervention-based Causal Discovery (ICD) and Causality-Inspired Interactive Policy Learning (CIIPL). Firstly, ICD aims to meticulously examine the causal relationships inherent in the interactive recommendation model, and formulates a novel Dynamic Causal Graph (DCG). The DCG is employed to generate the temporal conditional probability. Subsequently, the causal intervention is performed on the DCG, which serves to induce the temporal interventional probability. Afterwards, CIIPL utilizes the temporal conditional probability to estimate the cause-effect relationships between exposure bias and prediction scores during the phase of causality-inspired interactive training. It then employs the temporal interventional probability to alleviate the negative impact of exposure bias in the phase of causal intervention-based inference. Extensive experiments conducted on a series of public benchmarks substantiate the efficacy and efficiency of our proposed CIREC in mitigating exposure bias within the IRS.

Audio-driven talking face video generation has attracted increasing attention due to its huge industrial potential. Some previous methods focus on learning a direct mapping from audio to visual content. Despite progress, they often struggle with the ambiguity of the mapping process, leading to flawed results. An alternative strategy involves facial structural representations (e.g., facial landmarks) as intermediaries. This multi-stage approach better preserves the appearance details but suffers from error accumulation due to the independent optimization of different stages. Moreover, most previous methods rely on generative adversarial networks, prone to training instability and mode collapse. To address these challenges, our study proposes a novel landmark-based diffusion model for talking face generation, which leverages facial landmarks as intermediate representations while enabling end-to-end optimization. Specifically, we first establish the less ambiguous mapping from audio to landmark motion of lip and jaw. Then, we introduce an innovative conditioning module called TalkFormer to align the synthesized motion with the motion represented by landmarks via differentiable cross-attention, which enables end-to-end optimization for improved lip synchronization. Besides, TalkFormer employs implicit feature warping to align the reference image features with the target motion for preserving more appearance details. Extensive experiments demonstrate that our approach can synthesize high-fidelity and lip-synced talking face videos, preserving more subject appearance details from the reference image.

The acquisition of robotic skills via reinforcement learning (RL) is crucial for advancing embodied intelligence, but designing effective reward functions for complex tasks remains challenging. Recent methods using large language models (LLMs) can generate reward functions from language instructions, but they often produce overly goal-oriented rewards that neglect state exploration, causing robots to get stuck in local optima. Traditional RL addresses this by adding exploration bonuses, but these are typically generic and inefficient, wasting resources on exploring task-irrelevant areas. To address these limitations, we propose Policy-grounded Synergy of Reward Shaping and Exploration (PoRSE), a novel and unified framework that guides LLMs to generate task-aware reward functions while constructing an abstract affordance space for efficient exploration bonuses. Given the vast number of possible reward-bonus combinations, it is impractical to exhaustively train a policy from scratch for each configuration to identify the best one. Instead, PoRSE employs an in-policy-improvement grounding process, dynamically and continuously generating and filtering out reward-bonus pairs along the policy improvement process. This approach accelerates skill acquisition and fosters a mutually reinforcing relationship between reward shaping, exploration and policy enhancement through close feedback. Experiments show that PoRSE is highly effective, achieving significant improvement in average returns across all robotic tasks compared to previous state-of-the-art methods. It also achieves initial success in two highly challenging manipulation tasks, marking a significant breakthrough.

Achieving quadruped robot locomotion across diverse and dynamic terrains presents significant challenges, primarily due to the discrepancies between simulation environments and real-world conditions. Traditional sim-to-real transfer methods often rely on manual feature design or costly real-world fine-tuning. To address these limitations, this paper proposes the DreamTIP framework, which incorporates Task-Invariant Properties learning within the Dreamer world model architecture to enhance sim-to-real transfer capabilities. Guided by large language models, DreamTIP identifies and leverages Task-Invariant Properties, such as contact stability and terrain clearance, which exhibit robustness to dynamic variations and strong transferability across tasks. These properties are integrated into the world model as auxiliary prediction targets, enabling the policy to learn representations that are insensitive to underlying dynamic changes. Furthermore, an efficient adaptation strategy is designed, employing a mixed replay buffer and regularization constraints to rapidly calibrate to real-world dynamics while effectively mitigating representation collapse and catastrophic forgetting. Extensive experiments on complex terrains, including Stair, Climb, Tilt, and Crawl, demonstrate that DreamTIP significantly outperforms state-of-the-art baselines in both simulated and real-world environments. Our method achieves an average performance improvement of 28.1% across eight distinct simulated transfer tasks. In the real-world Climb task, the baseline method achieved only a 10% success rate, whereas our method attained a 100% success rate. These results indicate that incorporating Task-Invariant Properties into Dreamer learning offers a novel solution for achieving robust and transferable robot locomotion.

Synthesizing human motion from textual descriptions is essential for immersive digital applications, yet existing methods face a persistent trade-off between semantic fidelity and physical realism. Large language model (LLM)-based approaches can interpret diverse open-vocabulary instructions and compose high-level action plans, but they often generate motions that violate physical constraints. Physics-aware models improve realism through simulation or control, but they struggle with semantic complexity, fine-grained instructions, and novel concepts. To address this gap, we propose In-Context Model Predictive Generation (ICMPG), a framework that integrates language-model planning with inference-time physical feedback. ICMPG reformulates motion synthesis as a Model Predictive Control (MPC)-like process with two modules. The Context-Aware Motion Generation (CAMG) module uses an LLM as a planner to decompose textual commands and generate candidate motion sequences from motion tokens. The Model Predictive Generation (MPG) module evaluates these candidates through physical simulation and semantic alignment, estimates a composite reward, and selects the best sequence to guide subsequent generation steps. Unlike open-loop generation, this closed-loop refinement enables ICMPG to adapt motions to both the input semantics and the simulated physical environment without task-specific policy retraining. Extensive experiments across standard and zero-shot open-vocabulary settings show that ICMPG generalizes robustly to diverse commands and produces motions that are more physically plausible and semantically faithful than representative baselines on the evaluated benchmarks. The framework bridges semantic interpretation and physical simulation while remaining flexible enough to incorporate different LLM backbones, enabling more versatile and controllable text-driven motion synthesis.

Deep reinforcement learning (RL) has witnessed remarkable success in a wide range of control tasks. To overcome RL's notorious sample inefficiency, prior studies have explored data augmentation techniques leveraging collected transition data. However, these methods face challenges in synthesizing transitions adhering to the authentic environment dynamics, especially when the transition is high-dimensional and includes many redundant/irrelevant features to the task. In this article, we introduce continuous value assignment (CVA), an innovative optimization-level data augmentation approach that directly synthesizes novel training data in the state-action value space, effectively bypassing the need for explicit transition modeling. The key intuition of our method is that the transition plays an intermediate role in calculating the state-action value during optimization, and therefore directly augmenting the state-action value is more causally related to the optimization process. Specifically, our CVA combines parameterized value prediction and nonparametric value interpolation from neighboring states, resulting in doubly robust target values w.r.t. novel states and actions. Extensive experiments demonstrate CVA's substantial improvements in sample efficiency across complex continuous control tasks, surpassing several advanced baselines.

The ability to autonomously explore and resolve tasks with minimal human guidance is crucial for the self-development of embodied intelligence. Although reinforcement learning methods can largely ease human effort, it's challenging to design reward functions for real-world tasks, especially for high-dimensional robotic control, due to complex relationships among joints and tasks. Recent advancements large language models (LLMs) enable automatic reward function design. However, approaches evaluate reward functions by re-training policies from scratch placing an undue burden on the reward function, expecting it to be effective throughout the whole policy improvement process. We argue for a more practical strategy in robotic autonomy, focusing on refining existing policies with policy-dependent reward functions rather than a universal one. To this end, we propose a novel reward-policy co-evolution framework where the reward function and the learned policy benefit from each other's progressive on-the-fly improvements, resulting in more efficient and higher-performing skill acquisition. Specifically, the reward evolution process translates the robot's previous best reward function, descriptions of tasks and environment into text inputs. These inputs are used to query LLMs to generate a dynamic amount of reward function candidates, ensuring continuous improvement at each round of evolution. For policy evolution, our method generates new policy populations by hybridizing historically optimal and random policies. Through an improved Bayesian optimization, our approach efficiently and robustly identifies the most capable and plastic reward-policy combination, which then proceeds to the next round of co-evolution. Despite using less data, our approach demonstrates an average normalized improvement of 95.3\% across various high-dimensional robotic skill learning tasks.

Accurate inference of fine-grained traffic flow from coarse-grained one is an emerging yet crucial problem, which can help greatly reduce the number of the required traffic monitoring sensors for cost savings. In this work, we note that traffic flow has a high correlation with road network, which was either completely ignored or simply treated as an external factor in previous works. To facilitate this problem, we propose a novel road-aware traffic flow magnifier (RATFM) that explicitly exploits the prior knowledge of road networks to fully learn the road-aware spatial distribution of fine-grained traffic flow. Specifically, a multidirectional 1-D convolutional layer is first introduced to extract the semantic feature of the road network. Subsequently, we incorporate the road network feature and coarse-grained flow feature to regularize the short-range spatial distribution modeling of road-relative traffic flow. Furthermore, we take the road network feature as a query to capture the long-range spatial distribution of traffic flow with a transformer architecture. Benefiting from the road-aware inference mechanism, our method can generate high-quality fine-grained traffic flow maps. Extensive experiments on three real-world datasets show that the proposed RATFM outperforms state-of-the-art models under various scenarios. Our code and datasets are released at https://github.com/luimoli/RATFM.

Language-instructed robot manipulation has garnered significant interest due to the potential of learning from collected data. While the challenges in high-level perception and planning are continually addressed along the progress of general large pre-trained models, the low precision of low-level action estimation has emerged as the key limiting factor in manipulation performance. To this end, this paper introduces a novel robot manipulation framework, i.e., ActionSink, to pave the way toward precise action estimations in the field of learning-based robot manipulation. As the name suggests, ActionSink reformulates the actions of robots as action-caused optical flows from videos, called "action flow", in a self-supervised manner, which are then used to be retrieved and integrated to enhance the action estimation. Specifically, ActionSink incorporates two primary modules. The first module is a coarse-to-fine action flow matcher, which continuously refines the accuracy of action flow via iterative retrieval and denoising process. The second module is a dynamic action flow integrator, which employs a working memory pool that dynamically and efficiently manages the historical action flows that should be used to integrate to enhance the current action estimation. In this module, a multi-layer fusion module is proposed to integrate direct estimation and action flows from both the current and the working memory, achieving highly accurate action estimation through a series of estimation-integration processes. Our ActionSink framework outperformed prior SOTA on the LIBERO benchmark by a 7.9% success rate, and obtained nearly an 8% accuracy gain on the challenging long-horizon visual task LIBERO-Long.

Operating robots in open-ended scenarios with diverse tasks is a crucial research and application direction in robotics. While recent progress in natural language processing and large multimodal models has enhanced robots' ability to understand complex instructions, robot manipulation still faces the procedural skill dilemma and the declarative skill dilemma in open environments. Existing methods often compromise cognitive and executive capabilities. To address these challenges, in this paper, we propose RoBridge, a hierarchical intelligent architecture for general robotic manipulation. It consists of a high-level cognitive planner (HCP) based on a large-scale pre-trained vision-language model (VLM), an invariant operable representation (IOR) serving as a symbolic bridge, and a generalist embodied agent (GEA). RoBridge maintains the declarative skill of VLM and unleashes the procedural skill of reinforcement learning, effectively bridging the gap between cognition and execution. RoBridge demonstrates significant performance improvements over existing baselines, achieving a 75

Fine-tuning visual models has been widely shown promising performance on many downstream visual tasks. With the surprising development of pre-trained visual foundation models, visual tuning jumped out of the standard modus operandi that fine-tunes the whole pre-trained model or just the fully connected layer. Instead, recent advances can achieve superior performance than full-tuning the whole pre-trained parameters by updating far fewer parameters, enabling edge devices and downstream applications to reuse the increasingly large foundation models deployed on the cloud. With the aim of helping researchers get the full picture and future directions of visual tuning, this survey characterizes a large and thoughtful selection of recent works, providing a systematic and comprehensive overview of existing work and models. Specifically, it provides a detailed background of visual tuning and categorizes recent visual tuning techniques into five groups: fine-tuning, prompt tuning, adapter tuning, parameter tuning, and remapping tuning. Meanwhile, it offers some exciting research directions for prospective pre-training and various interactions in visual tuning.

Patients with triple-negative breast cancer (TNBC) frequently experience resistance to chemotherapy, leading to recurrence. The approach of optimizing anti-tumoral immunological effect is promising in overcoming such resistance, given the heterogeneity and lack of biomarkers in TNBC. In this study, we focused on YTHDF2, an N6-methyladenosine (m6A) RNA-reader protein, in macrophages, one of the most abundant intra-tumoral immune cells. Using single-cell sequencing and ex vivo experiments, we discovered that YTHDF2 significantly promotes pro-tumoral phenotype polarization of macrophages and is closely associated with down-regulated antigen-presentation signaling to other immune cells in TNBC. The in vitro deprivation of YTHDF2 favors anti-tumoral effect. Expressions of multiple transcription factors, especially SPI1, were consistently observed in YTHDF2-high macrophages, providing potential therapeutic targets for new strategies. In conclusion, YTHDF2 in macrophages appears to promote pro-tumoral effects while suppressing immune activity, indicating the treatment targeting YTHDF2 or its transcription factors could be a promising strategy for chemoresistant TNBC.

Language-guided robotic manipulation is a challenging task that requires an embodied agent to follow abstract user instructions to accomplish various complex manipulation tasks. Previous work generally maps instructions and visual perceptions directly to low-level executable actions, neglecting the modeling of critical waypoints (e.g., key states of “close to/grab/move up” in action trajectories) in manipulation tasks. To address this issue, we propose a PImitive-driVen waypOinT-aware world model for Robotic manipulation (PIVOT-R) that focuses solely on the prediction of task-relevant waypoints. Specifically, PIVOT-R consists of a Waypoint-aware World Model (WAWM) and a lightweight action prediction module. The former performs primitive action parsing and primitive-driven waypoint prediction, while the latter focuses on decoding low-level actions. Additionally, we also design an asynchronous hierarchical executor (AHE) for PIVOT-R, which can use different execution frequencies for different modules of the model, thereby helping the model reduce computational redundancy and improve model execution efficiency. Our PIVOT-R outperforms state-of-the-art (SoTA) open-source models on the SeaWave benchmark, achieving an average relative improvement of 19.45% across four levels of instruction tasks. Moreover, compared to the synchronously executed PIVOT-R, the execution efficiency of PIVOT-R with AHE is increased by 28-fold, with only a 2.9% drop in performance. These results provide compelling evidence that our PIVOT-R can significantly improve both the performance and efficiency of robotic manipulation.

Recent advancements utilizing large-scale video data for learning video generation models demonstrate significant potential in understanding complex physical dynamics. It suggests the feasibility of leveraging diverse robot trajectory data to develop a unified, dynamics-aware model to enhance robot manipulation. However, given the relatively small amount of available robot data, directly fitting data without considering the relationship between visual observations and actions could lead to suboptimal data utilization. To this end, we propose \textbf{VidMan} (\textbf{Vid}eo Diffusion for Robot \textbf{Man}ipulation), a novel framework that employs a two-stage training mechanism inspired by dual-process theory from neuroscience to enhance stability and improve data utilization efficiency. Specifically, in the first stage, VidMan is pre-trained on the Open X-Embodiment dataset (OXE) for predicting future visual trajectories in a video denoising diffusion manner, enabling the model to develop a long horizontal awareness of the environment's dynamics. In the second stage, a flexible yet effective layer-wise self-attention adapter is introduced to transform VidMan into an efficient inverse dynamics model that predicts action modulated by the implicit dynamics knowledge via parameter sharing. Our VidMan framework outperforms state-of-the-art baseline model GR-1 on the CALVIN benchmark, achieving a 11.7\% relative improvement, and demonstrates over 9\% precision gains on the OXE small-scale dataset. These results provide compelling evidence that world models can significantly enhance the precision of robot action prediction. Codes and models will be public.

Generating motion in line with text has attracted increasing attention nowadays. However, open-vocabulary human motion generation still remains touchless and undergoes the lack of diverse labeled data. The good news is that, recent studies of large foundation models (e.g., CLIP) have demonstrated superior performance on few/zero-shot image-text alignment, largely reducing the need for manually labeled data. In this paper, we take the advantage of CLIP for open-vocabulary 3D human motion generation in a zero-shot manner. Specifically, our model is composed of two stages, i.e., text2pose and pose2motion generations. For text2pose generation, to address the difficulty of optimization with direct supervision from CLIP, we propose to carve the versatile CLIP model into a slimmer but more specific model for aligning 3D poses and texts, via a novel pipeline distillation strategy. Optimizing with the distilled 3D pose-text model, we manage to concretize the text-pose knowledge of CLIP into a text2pose generator effectively and efficiently. As for pose2motion, drawing the inspiration of the advanced language model, we pretrain a transformer-based motion model, which makes up for the lack of motion dynamics of CLIP. After that, by formulating the condition poses as prompts, the motion generator can generate motions referring to the condition poses in a controllable and flexible manner.

Image blind denoising aims at removing the unknown noise from given images to improve the image’s visual quality. Current blind denoisers can be categorized into two classes, i.e., traditional and deep learning-based methods. Generally, deep learning-based methods are likely to outperform the traditional methods when the model of image noise is within the training noise distribution. However, when the image noise is unseen during the optimization process, traditional methods usually achieve better denoising performance. To address the generalization problem of deep learning-based denoiser, our paper proposed a general yet concise image blind denoising framework that can be applied to existing deep learning-based denoisers. Specifically, given noisy images, our method first extracts high-frequency signals from the images and synthesizes noise from the extracted signals using several noise generation techniques. In contrast to the existing conventional approaches that explicitly model the noise, our proposed framework advances in generating noise directly from the noisy images without explicitly noise modeling. By applying the generated noise to clean images to construct noisy-clean paired images, the deep learning-based denoisers can be optimized using the constructed data with the similar noise of the given noisy images. After optimization, the denoisers are capable of removing noise from the given noisy images. Extensive experiments and comprehensive ablations on both synthetic noisy image and real-world image denoising tasks demonstrate the superiority of our proposed framework over all the compared approaches.

Medical diagnosis assistant (MDA) aims to build an interactive diagnostic agent to sequentially inquire about symptoms for discriminating diseases. However, since the dialogue records for building a patient simulator are collected passively, the collected records might be deteriorated by some task-unrelated biases, such as the preference of the collectors. These biases might hinder the diagnostic agent to capture transportable knowledge from the simulator. This work identifies and resolves two representative non-causal biases, i.e., (i) default-answer bias and (ii) distributional inquiry bias. Specifically, Bias (i) originates from the patient simulator which tries to answer the unrecorded inquiries with some biased default answers. To eliminate this bias and improve upon a well-known causal inference technique, i.e., propensity score matching, we propose a novel propensity latent matching in building a patient simulator to effectively answer unrecorded inquiries; Bias (ii) inherently comes along with the passively collected data that the agent might learn by remembering what to inquire within the training data while not able to generalize to the out-of-distribution cases. To this end, we propose a progressive assurance agent, which includes the dual processes accounting for symptom inquiry and disease diagnosis respectively. The diagnosis process pictures the patient mentally and probabilistically by intervention to eliminate the effect of the inquiry behavior. And the inquiry process is driven by the diagnosis process to inquire about symptoms to enhance the diagnostic confidence which alters as the patient distribution changes. In this cooperative manner, our proposed agent can improve upon the out-of-distribution generalization significantly. Extensive experiments demonstrate that our framework achieves new state-of-the-art performance and possesses the advantage of transportability.

Video thumbnail plays an essential role in summarizing video content into a compact and concise image for users to browse efficiently. However, automatically generating attractive and informative video thumbnails remains an open problem due to the difficulty of formulating human aesthetic perception and the scarcity of paired training data. This work proposes a novel Human Perception-Centric Video Thumbnail Generation (HPCVTG) to address these challenges. Specifically, our framework first generates a set of thumbnails using a principle-based system, which conforms to established aesthetic and human perception principles, such as visual balance in the layout and avoiding overlapping elements. Then rather than designing from scratch, we ask human annotators to evaluate some of these thumbnails and select their preferred ones. A Transformer-based Variational Auto-Encoder (VAE) model is firstly pre-trained with Model-Agnostic Meta-Learning (MAML) and then fine-tuned on these human-selected thumbnails. The exploration of combining the MAML pre-training paradigm with human feedback in training can reduce human involvement and make the training process more efficient. Extensive experimental results show that our HPCVTG framework outperforms existing methods in objective and subjective evaluations, highlighting its potential to improve the user experience when browsing videos and inspire future research in human perception-centric content generation tasks. The code and dataset will be released via https://github.com/yangtao2019yt/HPCVTG.

The outbreak of COVID-19 reminds us of the urgency of reducing wild meat consumption (WMC). It is important to investigate why some people prefer wild meat. Using in-depth interviews (n = 19), group interviews (n = 9), and participant observation, this study conducted fieldwork in Nanxiong in South China in 2015-2016 and 2020, where WMC is considered a historical and cultural tradition. This study updates understandings of people's need for WMC in recent years and is likely the first attempt to apply social practice theory to investigate WMC, which provides an integrated and promising tool for theoretically summarizing the "motivations" in everyday life and dynamic relations behind this collective and conventional consumption. This paper offers new understandings of wild meat consumption through an exploration of how the related practices of preparing, cooking and tasting are changing. It explores materiality of taste and texture and the agentic capacity of wild meat. Additionally, the concept of practice memory is used to examine the development of WMC practice. In the context of COVID-19, some policy implications against WMC are provided.

Food taste should be considered in-place. This paper employs a regional analytical framework that is suggested by new regional geography to analyze relations between taste and place, and broadly, people and cuisine. We problematize fixed links between place and taste by asking how tastes are developed as rooted qualities of a place and its cuisine. Focusing on spicy taste as a cultural construct in China's context, climate and landscape characteristics, geographical proximity, and inhabitants' collective engagement are examined and identified as three factors that co-influence the construct of spicy taste, which we see as a vital shaper of the cuisine of Nanxiong in the mountainous north of China's southern Guangdong Province. We build a spatial regional hierarchy centered on spicy taste and find a correct and meaningful scale - the region of Guangdong Province. In highlighting the importance of considering the flavor principle in place, we call for more efforts to adopt a regional perspective in exploring multi-scalar and place-bounded food and taste issues.