Existing image inpainting methods have shown impressive completion results for low-resolution images. However, most of these algorithms fail at high resolutions and require powerful hardware, limiting their deployment on edge devices. Motivated by this, we propose the first baseline for REal-Time High-resolution image INpainting on Edge Devices (RETHINED) that is able to inpaint at ultra-high-resolution and can run in real-time ($\leq$ 30ms) in a wide variety of mobile devices. A simple, yet effective novel method formed by a lightweight Convolutional Neural Network (CNN) to recover structure, followed by a resolution-agnostic patch replacement mechanism to provide detailed texture. Specially our pipeline leverages the structural capacity of CNN and the high-level detail of patch-based methods, which is a key component for high-resolution image inpainting. To demonstrate the real application of our method, we conduct an extensive analysis on various mobile-friendly devices and demonstrate similar inpainting performance while being $\mathrm{100 \times faster}$ than existing state-of-the-art methods. Furthemore, we realease DF8K-Inpainting, the first free-form mask UHD inpainting dataset.

Vision-and-Language Navigation (VLN) systems often focus on either discrete (panoramic) or continuous (free-motion) paradigms alone, overlooking the complexities of human-populated, dynamic environments. We introduce a unified Human-Aware VLN (HA-VLN) benchmark that merges these paradigms under explicit social-awareness constraints. Our contributions include: 1. A standardized task definition that balances discrete-continuous navigation with personal-space requirements; 2. An enhanced human motion dataset (HAPS 2.0) and upgraded simulators capturing realistic multi-human interactions, outdoor contexts, and refined motion-language alignment; 3. Extensive benchmarking on 16,844 human-centric instructions, revealing how multi-human dynamics and partial observability pose substantial challenges for leading VLN agents; 4. Real-world robot tests validating sim-to-real transfer in crowded indoor spaces; and 5. A public leaderboard supporting transparent comparisons across discrete and continuous tasks. Empirical results show improved navigation success and fewer collisions when social context is integrated, underscoring the need for human-centric design. By releasing all datasets, simulators, agent code, and evaluation tools, we aim to advance safer, more capable, and socially responsible VLN research.

Natural language misinformation detection approaches have been, to date, largely dependent on sequence classification methods, producing opaque systems in which the reasons behind classification as misinformation are unclear. While an effort has been made in the area of automated fact-checking to propose explainable approaches to the problem, this is not the case for automated reason-checking systems. In this paper, we propose a new explainable framework for both factual and rational misinformation detection based on the theory of Argumentation Schemes and Critical Questions. For that purpose, we create and release NLAS-CQ, the first corpus combining 3,566 textbook-like natural language argumentation scheme instances and 4,687 corresponding answers to critical questions related to these arguments. On the basis of this corpus, we implement and validate our new framework which combines classification with question answering to analyse arguments in search of misinformation, and provides the explanations in form of critical questions to the human user.

We introduce MusicInfuser, an approach for generating high-quality dance videos that are synchronized to a specified music track. Rather than attempting to design and train a new multimodal audio-video model, we show how existing video diffusion models can be adapted to align with musical inputs by introducing lightweight music-video cross-attention and a low-rank adapter. Unlike prior work requiring motion capture data, our approach fine-tunes only on dance videos. MusicInfuser achieves high-quality music-driven video generation while preserving the flexibility and generative capabilities of the underlying models. We introduce an evaluation framework using Video-LLMs to assess multiple dimensions of dance generation quality. The project page and code are available at https://susunghong.github.io/MusicInfuser.

The CheckThat! lab aims to advance the development of innovative technologies designed to identify and counteract online disinformation and manipulation efforts across various languages and platforms. The first five editions focused on key tasks in the information verification pipeline, including check-worthiness, evidence retrieval and pairing, and verification. Since the 2023 edition, the lab has expanded its scope to address auxiliary tasks that support research and decision-making in verification. In the 2025 edition, the lab revisits core verification tasks while also considering auxiliary challenges. Task 1 focuses on the identification of subjectivity (a follow-up from CheckThat! 2024), Task 2 addresses claim normalization, Task 3 targets fact-checking numerical claims, and Task 4 explores scientific web discourse processing. These tasks present challenging classification and retrieval problems at both the document and span levels, including multilingual settings.

More than four years after its announcement and after two last-minute delays, the latest title in Ubisoft's historical fiction series Assassin's Creed will finally be released on Thursday. Set in Japan in 1579, a time of intense civil war dominated by the feudal lord Oda Nobunaga, it follows two characters navigating their way through the bloody chaos: a female shinobi named Fujibayashi Naoe, and Yasuke, an African slave turned samurai. Japan has been the series' most-requested setting for years, Ubisoft says. "I've been on [this] franchise for 16 years and I think every time we start a new game, Japan comes up and we ask, is this the time?" says executive producer Marc-Alexis Coté. "We've never pushed beyond the conception phase with Japan until this one." The game comes at a crucial time for Ubisoft after the disappointing performance of last year's titles Star Wars Outlaws, Skull and Bones and Prince of Persia: The Lost Crown, and the expensive closure of live service shooter XDefiant.

Hosted by Eleanor Drage and Kerry McInerney, The Good Robot is a podcast which explores the many complex intersections between gender, feminism and technology. To develop voice assistants like Siri and Alexa, companies spend years investigating what sounds like a human voice and what doesn't. But what we've ended up with is just one possibility of the kinds of voices that we could be interacting with. In this episode, we talked to sound engineer Frederik Juutilainen, and assistant professor at the University of Copenhagen, Stina Hasse Jørgensen, about their participation in [multi'vocal], an experimental research project that created an alternative voice assistant by asking people at a rock festival in Denmark to speak into a portable recording box. We talk about voice assistants' inability to stutter, lisp and code switch, and whether a voice can express multiple personalities, genders and ages.

We present PANDORA, a novel diffusion-based policy learning framework designed specifically for dexterous robotic piano performance. Our approach employs a conditional U-Net architecture enhanced with FiLM-based global conditioning, which iteratively denoises noisy action sequences into smooth, high-dimensional trajectories. To achieve precise key execution coupled with expressive musical performance, we design a composite reward function that integrates task-specific accuracy, audio fidelity, and high-level semantic feedback from a large language model (LLM) oracle. The LLM oracle assesses musical expressiveness and stylistic nuances, enabling dynamic, hand-specific reward adjustments. Further augmented by a residual inverse-kinematics refinement policy, PANDORA achieves state-of-the-art performance in the ROBOPIANIST environment, significantly outperforming baselines in both precision and expressiveness. Ablation studies validate the critical contributions of diffusion-based denoising and LLM-driven semantic feedback in enhancing robotic musicianship. Videos available at: https://taco-group.github.io/PANDORA

Multimodal Large Language Models (MLLMs) have shown impressive capabilities in reasoning, yet come with substantial computational cost, limiting their deployment in resource-constrained settings. Despite recent efforts on improving the efficiency of MLLMs, prior solutions fall short in responding to varying runtime conditions, in particular changing resource availability (e.g., contention due to the execution of other programs on the device). To bridge this gap, we introduce AdaLLaVA, an adaptive inference framework that learns to dynamically reconfigure operations in an MLLM during inference, accounting for the input data and a latency budget. We conduct extensive experiments across benchmarks involving question-answering, reasoning, and hallucination. Our results show that AdaLLaVA effectively adheres to input latency budget, achieving varying accuracy and latency tradeoffs at runtime. Further, we demonstrate that AdaLLaVA adapts to both input latency and content, can be integrated with token selection for enhanced efficiency, and generalizes across MLLMs. Our project webpage with code release is at https://zhuoyan-xu.github.io/ada-llava/.

Retrieval-Augmented Generation (RAG) has gained significant attention in recent years for its potential to enhance natural language understanding and generation by combining large-scale retrieval systems with generative models. RAG leverages external knowledge sources, such as documents, databases, or structured data, to improve model performance and generate more accurate and contextually relevant outputs. This survey aims to provide a comprehensive overview of RAG by examining its fundamental components, including retrieval mechanisms, generation processes, and the integration between the two. We discuss the key characteristics of RAG, such as its ability to augment generative models with dynamic external knowledge, and the challenges associated with aligning retrieved information with generative objectives. We also present a taxonomy that categorizes RAG methods, ranging from basic retrieval-augmented approaches to more advanced models incorporating multi-modal data and reasoning capabilities. Additionally, we review the evaluation benchmarks and datasets commonly used to assess RAG systems, along with a detailed exploration of its applications in fields such as question answering, summarization, and information retrieval. Finally, we highlight emerging research directions and opportunities for improving RAG systems, such as enhanced retrieval efficiency, model interpretability, and domain-specific adaptations. This paper concludes by outlining the prospects for RAG in addressing real-world challenges and its potential to drive further advancements in natural language processing.