金磊 发自 杭州 量子位 | 公众号 QbitAI 天下苦机器人看不清 透明 和 反光 物体久矣。 毕竟就连小动物甚至人，有时候一个不小心，都会搞笑地撞到干净的玻璃门…… 不仅如此，若是让机器人拿起 透明的玻璃杯 、 反光的不锈钢 物体，他们也会经常出现"突然瞎了"的情况。 这一切的问题，正是出在了机器人的眼睛—— 深度相机 。 因为无论是基于结构光还是双目立体视觉的深度相机，它们的工作原理都是依赖物体表面对光线的稳定反射。 而透明材质会让光线直接穿透，高反光材质则会将光线漫反射到四面八方，导致传感器无法接收到有效的回波信号，从而产生大量缺失或错 误的深度值。 对比一下我们人类看到的场景和机器人眼中的场景，就一目了然了： 毫不夸张地说，这类让机器人睁眼瞎的问题，一直是阻碍它们安全地走进家庭、商场和医院等场景的 Big Big Big Problem！ 但现在，随着一项新技术的提出，机器人的眼疾终于算是被治好了—— 蚂蚁集团的具身智能公司 蚂蚁灵波 （RobbyAnt），开源了全球看得最清楚的深度视觉模型， LingBot-Depth 。 同样是上面两个场景，我们直接来看下在LingBot-Depth加持下的效 ...

Core Insights - Ant Group's subsidiary Lingbo Technology has officially open-sourced the high-precision spatial perception model LingBot-Depth, focusing on enhancing environmental depth perception and 3D spatial understanding capabilities [1] - The model is developed using chip-level raw data from the Gemini330 series binocular 3D cameras provided by Orbbec, aiming to deliver more accurate and reliable 3D vision solutions for robots and autonomous driving [1] - This collaboration signifies a breakthrough in the critical technology of enabling machines to "see the 3D world," marking Lingbo Technology's first major progress announcement in the field of embodied intelligence since its appearance at the 2025 Bund Conference six months ago [1] - Following the announcement, Orbbec's stock price rose against market trends, with an intraday increase of over 6%, indicating a significant rise in market attention [1]

Core Insights - Ant Group's Lingbo Technology has made significant advancements in spatial intelligence by open-sourcing the high-precision spatial perception model LingBot-Depth, which enhances depth perception and 3D spatial understanding for robots and autonomous vehicles [1] Group 1: Model Performance - LingBot-Depth demonstrates a generational advantage in authoritative benchmark evaluations, reducing relative error (REL) by over 70% compared to mainstream models like PromptDA and PriorDA in indoor scenes, and achieving a 47% reduction in RMSE error in challenging sparse SfM tasks [1] - The model excels in handling transparent and reflective objects, which are common in household and industrial environments, overcoming limitations faced by traditional depth cameras [1][2] Group 2: Technology and Innovation - The "Masked Depth Modeling" (MDM) technology developed by Lingbo Technology allows the model to infer and complete missing depth data by integrating texture, contours, and contextual information from RGB images, resulting in clearer and more complete 3D depth maps [2] - LingBot-Depth has been certified by the Oubo Zhongguang Depth Vision Laboratory, achieving industry-leading levels in accuracy, stability, and adaptability to complex scenes [2] Group 3: Data and Collaboration - The model's superiority is attributed to a vast dataset, with approximately 10 million raw samples and 2 million high-value depth pair data used for training, which will soon be open-sourced to accelerate community efforts in tackling complex spatial perception challenges [3] - Ant Group's Lingbo Technology has reached a strategic cooperation intention with Oubo Zhongguang to launch a new generation of depth cameras based on LingBot-Depth's capabilities [3]

LingBot-Depth 的优异性来源于海量真实场景数据。灵波科技采集约 1000 万份原始样本，提炼出 200 万 组高价值深度配对数据用于训练，支撑模型在极端环境下的泛化能力。这一核心数据资产（包括 2M 真 实世界深度数据和 1M 仿真数据）将于近期开源，推动社区更快攻克复杂场景空间感知难题。 空间智能迎来重要开源进展。1月27日，蚂蚁集团旗下具身智能公司灵波科技宣布开源高精度空间感知 模型LingBot-Depth。 该模型基于奥比中光 Gemini 330 系列双目 3D 相机提供的芯片级原始数据，专注于提升环境深度感知与 三维空间理解能力，旨在为机器人、自动驾驶汽车等智能终端赋予更精准、更可靠的三维视觉，在"看 清楚"三维世界这一行业关键难题上取得重要突破。这也是蚂蚁灵波科技在2025外滩大会后首次亮相 后，时隔半年在具身智能技术基座方向公布重要成果。 在NYUv2、ETH3D等权威基准评测中，LingBot-Depth展现出代际级优势：相比业界主流的 PromptDA 与PriorDA，其在室内场景的相对误差（REL）降低超过70%，在挑战性的稀疏SfM 任务中RMSE误差降 低约47%。 在 ...

Core Insights - Ant Group's subsidiary Lingbo Technology has announced the open-source release of a high-precision spatial perception model called LingBot-Depth [1][2] - The model is based on raw chip-level data from the Orbbec Gemini 330 series dual-camera 3D system, focusing on enhancing environmental depth perception and three-dimensional spatial understanding [1][2] - The aim is to provide more accurate and reliable 3D vision for intelligent terminals such as robots and autonomous vehicles, addressing a critical industry challenge in "seeing" the three-dimensional world [1][2] - This announcement marks the first significant achievement in embodied intelligence technology by Lingbo Technology since its debut at the 2025 Bund Conference six months ago [1][2]

B F 10 in A i 11 B 的座報 给地 0 a 千年宋韵 Fristeri n KAT 目# 康房体验 a 2026，岁在丙午，骏马"宋"福而至。 马，是刻在中国人文化基因里的奋进图腾。它是"车如流水马如龙"的市井繁华，是"快马加鞭未下 鞍"的奋进姿态，更是"春风得意马蹄疾"的美好期许。从旅游标志"马踏飞燕"的天马行空，到杭州自 古"东南形胜"所孕育的开拓精神，"马"的内涵与这座城市的脉搏深度共鸣。当马的精神遇见宋的风雅， 一场专属于杭州的春节仪式，便在千年传承中焕发新的生机——这不仅是一场庆典，更是一次古今交融 的文化复兴，一条连接传统与现代的情感纽带。 1月23日晚，由杭州市文化广电旅游局、杭州市上城区人民政府共同主办的"百骏迎新·2026'宋'福杭 州年"启动仪式，在湖滨步行街西子廊桥举行。伴随着满载福气的"马灯福气船"缓缓启航，以及"马上钱 潮"的铜艺巨制惊艳亮相，一幅生动鲜活的杭州新春文旅"百骏图"徐徐展开——三大事件、五大产品、 六大场景、百项活动，在为期两个月的时间里，将传统年味、宋韵美学与现代文旅创新有机融合，开启 一场贯穿整个春节的沉浸式文化盛宴，邀市民游客共赴一场跨越时空、充满 ...

©️深响原创 · 作者｜何理 毫无意外，AI成了互联网大厂们的必争赛点。 豆包投流高歌猛进、阿里先推夸克后捧千问、腾讯元宝怒撒10亿红包、百度跟进5亿……不管是研发投入、技术积累，还是广告投流、品牌营销、红包撒 钱，各家都是"饱和式"投入，欲在新一轮竞争中拔得头筹。 而这波AI浪潮里，一个老玩家的新面孔让人不得不提高注意力，蚂蚁阿福后程发力，新版上线不到1个月，最新月活用户已突破3000万。这印证了市场对 其的期待——过去五年，蚂蚁不再站在聚光灯下，但无论它的体量、所处的关键入口位置，还是技术积累，都明示着它不会、也不能在AI的未来战役中 缺席。 一边是蚂蚁阿福和灵光的声量骤起，一边是蚂蚁集团CEO 韩歆毅在晚点专访中首次完整披露公司未来十年战略方向。诸多迹象表明，蚂蚁已经重回战场 了。 "两朵花"， 重新理解蚂蚁的新战略 可以想象，蚂蚁在沉寂的五年里一定憋着一口气。在那段时间里，它并非什么都没做，各种业务创新一直在推进，但始终不足以成为支撑蚂蚁下一个十年 的"核心创新"。 韩歆毅在采访中坦言他反复思考的问题是——蚂蚁下一个真正站得住的创新是什么？ 转折点出现在2025 年春节前后，DeepSeek带来的冲击 ...

Core Viewpoint - The CEO of Tianghu Robotics predicts that the year 2026 will mark a critical point in the delivery battle within the embodied intelligence industry, where companies failing to deliver on time and quality will be eliminated [1][10]. Company Overview - Tianghu Robotics, established five years ago, specializes in the production of robotic joint modules and assembly of robots, with plans to expand its annual production capacity from 30,000 to 100,000 joint modules by 2025 and to assemble 3,000 humanoid robots annually [1][2]. - The company has attracted investments from notable institutions such as Ant Group, China Mobile, and SenseTime due to its production capacity [2]. Industry Challenges - The industry is currently facing challenges related to mass production and delivery, with many companies struggling to fulfill orders despite high demand [2][12]. - The complexity of the entire delivery chain, which includes design, procurement, assembly, testing, and after-sales service, is often underestimated, leading to significant operational difficulties [12][10]. Delivery and Trust Issues - The CEO emphasizes that the ability to refuse orders is more critical than merely acquiring them, as over-promising can lead to a trust crisis [4][16]. - Tianghu has experienced delays and quality issues with suppliers, which have impacted its delivery timelines and customer trust [9][18]. Future Outlook - The company anticipates that by 2026, firms unable to deliver will face severe cash flow issues, potentially leading to their demise [10][12]. - The CEO believes that the market for humanoid robots will see significant growth, with potential shipments reaching one million units between 2026 and 2028, followed by a more substantial industry explosion around 2030 to 2035 [27].

Core Insights - OpenEvidence has rapidly gained traction in the medical field, achieving a valuation of $12 billion and annual revenue exceeding $150 million within just four years of its establishment [1] - The company addresses a critical gap in the medical industry by providing a free tool for doctors that significantly reduces the time needed to access reliable medical information [4][5] - OpenEvidence's business model revolves around monetizing the attention of healthcare professionals and providing targeted advertising for pharmaceutical companies [7][9][10] Group 1: OpenEvidence's Rise - OpenEvidence has become the primary entry point for doctors by effectively addressing the overwhelming volume of medical knowledge and the limitations of traditional databases [2][3] - The platform utilizes a retrieval-augmented generation (RAG) approach, allowing doctors to obtain accurate information in just three seconds, thus enhancing decision-making efficiency [4] - The company has achieved viral growth, with monthly active users reaching 400,000 and covering approximately 34% of practicing physicians in the U.S. [5] Group 2: Revenue Generation - OpenEvidence generates revenue by providing targeted advertising to pharmaceutical companies during critical decision-making moments for doctors [8][9] - The platform's ability to deliver compliant and relevant advertising content has made it an attractive option for drug companies looking to reach physicians effectively [10][12] - Additionally, OpenEvidence sells its core capabilities as APIs to hospitals and medical schools, further diversifying its revenue streams [11] Group 3: Challenges for Chinese Competitors - Chinese companies face significant challenges in replicating OpenEvidence's success due to data integration difficulties and the lack of open access to authoritative medical databases [15][16] - Trust issues arise in China regarding pharmaceutical advertising alongside clinical decision tools, making it difficult for companies to monetize similar models [17][18] - The high workload of Chinese doctors limits their ability to engage with tools like OpenEvidence, necessitating a more practical approach tailored to local conditions [19][20] Group 4: Competitive Landscape - JD Health focuses on a model that combines tools, supply chain, and services, but faces trust issues due to potential biases in its recommendations [23][24] - Alibaba Health aims to develop a comprehensive medical operating system but struggles with the transactional aspect of its services [25][26] - Ant Group's approach with its AI tool "Afu" seeks to integrate deeply into the medical workflow, potentially offering a more complex but rewarding business model [27][28] Group 5: Future Outlook - The medical AI market in China is expected to diversify, with different players targeting various segments, such as serious medical scenarios and primary care [29] - The key lesson from OpenEvidence for Chinese companies is to effectively use free tools to capture high-value users and monetize their needs [29]

Core Viewpoint - Tencent is launching a Spring Festival event on February 1, distributing a total of 1 billion cash red envelopes through the Yuanbao app, with individual envelopes worth up to 10,000 yuan [1][4]. Group 1: Tencent's Spring Festival Activity - The event will feature various forms of red envelopes, including a 10,000 yuan "Little Horse Card" and cash red envelopes, with users able to earn more through daily tasks [3][4]. - Users who successfully reserve their participation will receive an additional 10 chances to win on the event day [1][4]. - The cash can be directly withdrawn to WeChat from Yuanbao [1]. Group 2: AI Integration and Market Position - Tencent is reportedly preparing to introduce new AI features in Yuanbao, with a new tab labeled "Distribute" being tested [5]. - Yuanbao has seen significant growth since integrating with DeepSeek, with user engagement increasing over 100 times since early last year, positioning it among the top three native AI applications in China [8]. Group 3: Competitive Landscape - Other companies are also launching similar initiatives, with Baidu announcing a 500 million yuan red envelope campaign for the 2026 Spring Festival, emphasizing AI interaction [8][12]. - The trend of using AI as a core feature for red envelope distribution reflects a shift in the industry, as companies aim to enhance user engagement beyond mere cash incentives [13]. - The competitive landscape for Spring Festival red envelopes has evolved, with major players like Alibaba and ByteDance also participating, indicating a strategic focus on user retention and experience [13].