根据彭思达个人学术主页，至少与理想合作了3篇论文，分别是： 2026年1月6日发布的InfiniDepth: Arbitrary-Resolution and Fine-Grained Depth Estimation with Neural Implicit Fields 彭思达为通讯作者 今天本来打算发这篇论文的解读文的，但因为时间来不及写高质量解读文就明天再发了。 2024年12月17日第一版，2025年8月26日第三版的StreetCrafter: Street View Synthesis with Controllable Video Diffusion Models 彭思达为通讯作者。 谷歌学术主页链接： https://scholar.google.com/citations?user=l9NCksYAAAAJ&hl=en GitHub上有2.3K关注者，可以看到多篇高引用文章是公布了开源代码，并获得了高星。 | | Pinned | | | --- | --- | --- | | | 2 zju3dv/neuralbody (Public | 2 zju3dv/animatable_nerf ...

在 计算机图形学、 三维视觉 、虚拟人、XR 领域 ，SIGGRAPH 是毫无争议的 "天花板级会议"。 SIGGRAPH Asia 作为 SIGGRAPH 系列两大主会之一，每年只接 收全球最顶尖研究团队的成果稿件，代表着学术与工业界的 最高研究水平与最前沿技术趋势 。 我们是淘宝技术 - Meta 技术团队，在 3D、XR、3D 真人数字人和三维重建等方向拥有深厚的技术积累和业务沉淀，我们自研了专业的多视角拍摄影棚，在今年 CVPR 2025 会议上作为 Highlight Paper 发表了 TaoAvatar ，并在淘宝未来旗舰店中实现了业内首个 3D 真人导购体验，下面视频展示了杭州西溪园区 C 区淘宝未来 旗舰店的精彩瞬间，欢迎大家到来访园区进行体验。 今年我们团队迎来另一个重要里程碑：我们撰写的针对移动端的高保真实时 3D 数字人重建与渲染系统论文 首次登录了国际顶级计算机图形学会议 SIGGRAPH Asia ！这是我们技术实力的一次正式 "官宣"，也是我们在 3D/XR 方向长期投入的阶段性成果展示。 我们研发的基于手机单目视频生成高保真且可实时驱动的 3D 数字人的系统名叫 HRM²Ava ...

机器之心报道 机器之心编辑部 一年一度的英伟达奖学金出炉了。 二十五年来，英伟达研究生奖学金计划（NVIDIA Graduate Fellowship Program）一直为研究生提供与英伟达技术相关的杰出工作支持。 今天，该计划宣布了 2026 年度的 10 位博士生获奖者，他们每人将获得最高 6 万美元的资助，以支持他们在涵盖计算创新所有领域的各项研究。 他们的研究工作聚焦于加速计算的前沿领域，包括了自主系统、计算机体系结构、计算机图形学、深度学习、编程系统、机器人技术和安全。 本年度的 10 位获奖者中有 8 位华人 。去年有 7 位华人博士生 入选，包括了上交、中科大、浙大校友。 接下来，我们一起了解下本年度获奖者的信息。 Jiageng Mao 南加州大学，获奖理由：利用来自互联网规模数据的各种先验知识解决复杂的物理人工智能问题，从而为现实世界中的具身智能体实现稳健、可推广的智能。 资料显示，Jiageng Mao 是南加州大学博士生，研究方向是物理人工智能，目标是通过开发机器人、计算机视觉和自然语言处理等领域的算法，将人工智能应用于 现实世界。据了解，他对直观物理学、大型视觉 - 语言（- 动作） ...

Core Insights - The article discusses the rising importance of 3DGS (3D Geometry Synthesis) technology in various fields, particularly in autonomous driving, healthcare, virtual reality, and gaming [2][4] - A comprehensive learning roadmap for 3DGS has been developed to address the industry's need for effective training in scene reconstruction and world modeling [4][6] Course Overview - The course titled "3DGS Theory and Algorithm Practical Tutorial" aims to provide a detailed understanding of 3DGS algorithms, covering both theoretical foundations and practical applications [6][10] - The course is designed in six chapters, starting from basic knowledge to advanced research directions in 3DGS [10][11] Chapter Summaries - **Chapter 1: Background Knowledge** Introduces foundational concepts in computer graphics, including implicit and explicit representations of 3D space, rendering pipelines, and tools like SuperSplat and COLMAP [10][11] - **Chapter 2: Principles and Algorithms** Focuses on the core principles of 3DGS, including dynamic and surface reconstruction, and introduces the 3DGRUT framework for practical learning [11][12] - **Chapter 3: 3DGS in Autonomous Driving** Highlights key works in the field, such as Street Gaussian and OmniRe, and utilizes DriveStudio for practical applications [12][13] - **Chapter 4: Important Research Directions** Discusses significant research areas like COLMAP extensions and depth estimation, emphasizing their relevance to both industry and academia [13][14] - **Chapter 5: Feed-Forward 3DGS** Explores the development and principles of feed-forward 3DGS, including recent algorithms like AnySplat and WorldSplat [14][15] - **Chapter 6: Q&A Discussion** Provides a platform for participants to discuss industry pain points and job demands related to 3DGS [15] Target Audience and Learning Outcomes - The course is aimed at individuals with a background in computer graphics, visual reconstruction, and programming, particularly those interested in pursuing careers in the 3DGS field [19][17] - Participants will gain comprehensive knowledge of 3DGS theory, algorithm development frameworks, and opportunities for networking with industry professionals [19][17]

Core Viewpoint - The research team led by Professor Huang Hui from Shenzhen University has developed a new generation of intelligent monitoring system for geological disasters, which integrates computer vision, deep learning, and cloud-edge-end collaborative technology, transforming traditional point-based monitoring into comprehensive and intelligent monitoring [1][3]. Group 1: Traditional Monitoring Limitations - Traditional geological disaster monitoring methods rely heavily on embedded sensors and manual inspections, which have significant limitations [3]. - Sensors can only monitor preset points and cannot cover entire risk areas, while manual inspections are constrained by weather and terrain, making many dangerous areas inaccessible [3]. Group 2: Technological Innovations - The team proposed a core graphic information "cloud-edge-end" collaborative processing technology, achieving a transition from point monitoring to comprehensive prevention [3]. - The system utilizes a combination of computer graphics, computer vision, and deep learning, with breakthroughs in three key technical areas: effective capture of abnormal movements in monitored areas, over 85% accuracy in identifying rockfall events, and high-precision measurement of target displacement [3]. Group 3: Application and Impact - The system has demonstrated its application value in various scenarios, including 24-hour monitoring of tunnel entrances and high slope sections on mountain roads, rockfall disaster warnings for railways, stability monitoring in open-pit mining, and ensuring the safety of slopes in water conservancy projects [5]. - It has been implemented in Shenzhen's Jiangangshan Park, providing continuous monitoring and alarm for dangerous rocks and rockfalls [5]. - The monitoring device is equipped with a large-capacity solar power system for uninterrupted operation, showcasing strong environmental adaptability and energy self-sufficiency [5]. - The system captures minute changes in rock formations using high-resolution cameras and analyzes data in real-time with built-in intelligent algorithms, triggering multi-level alerts and uploading data to a cloud management platform via 4G/5G networks [5]. - This technology marks a shift from passive waiting to proactive prediction in geological disaster monitoring and early warning, entering a new phase of "full-domain perception, intelligent deduction, and precise warning" [5].

Core Viewpoint - The article discusses the development of Sketch3DVE, a novel method for 3D scene video editing that allows users to manipulate videos using simple sketches, enhancing creativity and personalization in video content creation [3][22]. Part 1: Background - Recent advancements in video generation models have significantly improved text-to-video and image-to-video generation, with a focus on precise control over camera trajectories due to its important application prospects [6]. - Existing methods for video editing are categorized into two types: one directly uses camera parameters as model inputs, while the other constructs explicit 3D representations from single images to render new perspective images [8][9]. - Despite these advancements, editing real videos with significant camera motion remains a challenge, as video editing requires maintaining original motion patterns and local features while synthesizing new content [8][9]. Part 2: Algorithm Principles - Users begin by selecting the first frame of a 3D scene video, marking the editing area with a mask and drawing a sketch to specify the geometry of new objects [12]. - The system employs the MagicQuill image editing algorithm to process the first frame, generating the edited result, and utilizes the DUSt3R algorithm for 3D reconstruction to analyze the entire input video [13]. - A 3D mask propagation algorithm is designed to accurately transfer the mask from the first frame to subsequent frames, ensuring consistency across different perspectives [14]. - The final video generation model integrates edited images, multi-view videos, and original input videos to produce a scene-edited video with precise 3D consistency [14]. Part 3: Effect Demonstration - The method allows users to create high-quality 3D scene video edits, enabling operations such as adding, removing, and replacing objects while maintaining good 3D consistency [16]. - The approach can handle complex scenarios involving shadows and reflections, producing reasonable editing results due to training on real video datasets [17]. - Users can also edit the first frame using image completion methods, demonstrating the versatility of the system in generating realistic 3D scene video edits [19]. - Sketch3DVE offers an effective solution to traditional model insertion challenges, allowing for personalized 3D object generation and high-fidelity scene video editing without requiring extensive expertise [22].

Core Insights - The article emphasizes the rapid transformation of the world through artificial intelligence and programming technologies, highlighting the significance of Auckland University's computer science undergraduate program as a platform for students passionate about these fields [1]. Group 1: Program Advantages - Auckland University's computer science program boasts exceptional academic resources and a strong faculty, with the department recognized internationally for its research in artificial intelligence, data science, and cybersecurity [3]. - The faculty comprises professors from around the globe who have made significant academic contributions and maintain close collaborations with major tech companies like Google and Microsoft, integrating the latest industry trends into the curriculum [3]. - The university provides advanced learning resources, including high-performance computing clusters and virtual reality equipment, facilitating complex programming experiments and AI project development [3]. - Partnerships with numerous tech companies offer students internship and employment opportunities, allowing them to engage with real-world business projects during their studies [3]. Group 2: Application Requirements - Applicants to the computer science undergraduate program must meet specific academic and language criteria, with international students typically required to achieve an average high school score of over 80%, particularly excelling in mathematics and physics [4]. - For Chinese students, the Gaokao score is a critical reference, generally requiring a score above the provincial first-tier line; alternative qualifications like A-Level or IB scores are also accepted [4]. - Language proficiency is essential, with a minimum IELTS score of 6.5 (no individual score below 6.0) or a TOEFL score of 90 (with writing no less than 21) required for admission [4]. Group 3: Curriculum Content - The curriculum is diverse and designed to build a solid theoretical foundation and practical innovation skills, starting with introductory courses in computer science, programming basics (Python and Java), and discrete mathematics in the first year [6]. - As students progress, they encounter more specialized courses such as data structures and algorithms, computer systems principles, and database systems, deepening their understanding of computer science fundamentals [6]. - Elective courses in artificial intelligence, machine learning, computer graphics, and cybersecurity allow students to explore cutting-edge areas of interest, while project-based courses enable teamwork and problem-solving through real programming projects [6].