扫码与新民体育 一起探访花滑转播新科技 打好辅助 与短道速滑、速度滑冰这些毫秒必争的计时项目不同，花样滑冰从来不是"看谁更快"，技术完成度、艺 术表现力的层层评判，让它始终伴随着判罚争议。而全新三维追踪系统，正是为了让这份"主观"多一份 客观支撑。 过去，裁判只能凭经验判断选手转体是否完整、落冰是否干净，外刃还是内刃的争议、不足周的细微偏 差，往往藏在肉眼难辨的瞬间。而如今，计算机视觉技术甚至能清晰捕捉冰刀刀刃的角度与落点位置。 正如工作人员所言，这一次，技术让裁判"看到了过去看不到的细节"。 值得一提的是，在米兰冬奥会上，这套量化系统始终保持着"克制"。阿兰在采访中反复强调，这些精准 的运动数据并不会直接参与裁判打分，而主要用于电视转播和赛后分析。 毫厘毕现 对观众而言，这场"量化革命"让花样滑冰变得更易懂。电视转播中，4K超高清图形技术将跳跃数据与 画面完美融合，一段优美的跳跃不再只是"看上去很高""转得很快"，而是有了具体的数值参考，观众能 直观感受到选手动作的精妙之处。 冰刀在冰面上拉出一道弧线，哈萨克斯坦选手萨多洛夫腾空而起，四周跳如流星破空，腾跃、旋转、落 定，掌声如潮。慢动作回放中，萨多洛夫的身 ...

中石化（大连）石油化工研究院有限公司，成立于2022年，位于大连市，是一家以从事专业技术服务业 为主的企业。企业注册资本338459.08万人民币。通过天眼查大数据分析，中石化（大连）石油化工研 究院有限公司共对外投资了1家企业，参与招投标项目2994次，专利信息5000条，此外企业还拥有行政 许可65个。 国家知识产权局信息显示，中国石油化工股份有限公司;中石化（大连）石油化工研究院有限公司申请 一项名为"废弃光伏板上贵金属识别方法、装置、电子设备及存储介质"的专利，公开号 CN121438279A，申请日期为2024年7月。 专利摘要显示，本发明提供一种废弃光伏板上贵金属识别方法、装置、电子设备及存储介质，涉及计算 机视觉技术领域，该方法首先获取废弃光伏板的待识别图像；然后将待识别图像输入至贵金属识别模 型，得到贵金属识别模型输出的废弃光伏板上的贵金属类别。该方法中采用的贵金属识别模型通过注意 力机制，采用金字塔卷积神经网络，可以准确提取出待识别图像中贵金属相关的目标特征图，并利用目 标特征图，确定贵金属类别。该方法通过贵金属识别模型，可以实现对贵金属类别的自动识别，提高识 别效率，提升识别结果的准确性， ...

Group 1 - The core viewpoint of the news is that Geling Deep Vision's stock performance and financial metrics indicate a mixed outlook, with significant revenue growth but ongoing net losses [1][2]. - On January 22, Geling Deep Vision's stock fell by 0.22%, with a trading volume of 131 million yuan. The financing buy amount was 13.59 million yuan, while the financing repayment was 12.26 million yuan, resulting in a net financing purchase of 1.33 million yuan [1]. - As of January 22, the total margin balance for Geling Deep Vision was 183 million yuan, accounting for 3.98% of its market capitalization, indicating a high level of financing compared to the past year [1]. Group 2 - As of September 30, the number of shareholders for Geling Deep Vision was 15,200, a decrease of 2.49% from the previous period, while the average circulating shares per person increased by 40.76% to 17,075 shares [2]. - For the period from January to September 2025, Geling Deep Vision reported revenue of 94.43 million yuan, representing a year-on-year growth of 55.33%. However, the net profit attributable to shareholders was a loss of 127 million yuan, although this was an improvement of 7.47% year-on-year [2]. - Among institutional holdings, as of September 30, 2025, the top ten circulating shareholders included Huabao Zhongzheng Financial Technology Theme ETF, which held 3.52 million shares, an increase of 1.73 million shares from the previous period, and Yifangda Guozheng Robotics Industry ETF, which was a new shareholder with 3.39 million shares [2].

1月16日，格灵深瞳跌5.07%，截至发稿，报17.80元/股，成交1.34亿元，换手率2.82%，总市值46.10亿 元。 华宝中证金融科技主题ETF（159851）基金经理为陈建华、曹旭辰。 截至发稿，陈建华累计任职时间13年29天，现任基金资产总规模277.02亿元，任职期间最佳基金回报 184.66%， 任职期间最差基金回报-49.65%。 责任编辑：小浪快报 数据显示，华宝基金旗下1只基金位居格灵深瞳十大流通股东。华宝中证金融科技主题ETF（159851） 三季度增持173.23万股，持有股数351.76万股，占流通股的比例为1.36%。根据测算，今日浮亏损失约 334.17万元。 华宝中证金融科技主题ETF（159851）成立日期2021年3月4日，最新规模123.19亿。今年以来收益 10.08%，同类排名1034/5531；近一年收益36.11%，同类排名2228/4215；成立以来收益86.28%。 曹旭辰累计任职时间240天，现任基金资产总规模351.82亿元，任职期间最佳基金回报126.35%， 任职 期间最差基金回报-4.65%。 资料显示，北京格灵深瞳信息技术股份有限公司位于北京市海 ...

Group 1 - The State Intellectual Property Office of China shows that Shandong Qilu Yidian Media Co., Ltd. and Shandong Digital Culture Group Co., Ltd. have applied for a patent titled "A Collaborative Distillation and Quantization Compression Method for Image Semantic Representation," with publication number CN121305300A and application date of October 2025 [1] - The patent application falls under the field of computer vision technology, specifically involving a collaborative distillation and quantization compression method that deeply couples vector quantization and knowledge distillation through bidirectional information interaction for collaborative optimization, achieving a high compression ratio while maximizing the quality of the model's semantic representation [1] - Shandong Qilu Yidian Media Co., Ltd. was established in 2014, located in Jinan, with a registered capital of 30 million RMB, and has invested in one company, participated in 579 bidding projects, and holds 59 trademark records and 52 patent records [1] - Shandong Digital Culture Group Co., Ltd. was established in 2013, also located in Jinan, with a registered capital of 110 million RMB, and has invested in six companies, participated in 16 bidding projects, and holds eight patent records [2]

Core Insights - The rapid development of the cross-border e-commerce industry has made visual content a key factor in attracting consumers and enhancing conversion rates [1] - AI-generated video technology effectively addresses the content production challenges faced by cross-border merchants in multi-market and multi-language environments [1] - The technology has been widely adopted on platforms like Amazon, eBay, and TikTok Shop, becoming an important tool for improving overseas marketing effectiveness [1] Group 1: AI Video Generation Technology - The core of AI-generated video technology lies in combining generative artificial intelligence with computer vision to dynamically reconstruct static images [1] - Deep learning algorithms enable the automatic generation of product displays, model demonstrations, and usage scenario videos, supporting multi-language subtitles and localized elements [1] - This technology significantly reduces production costs and allows for rapid content iteration across cultural markets [1] Group 2: Key Tools in the Market - Keevx focuses on providing efficient video generation solutions for cross-border e-commerce, quickly generating virtual model display videos suitable for product detail pages, platform ads, and social media marketing [1] - Runway Gen-2 is recognized for its strong video generation capabilities and multi-scene adaptability, supporting high-quality short video creation from static images [3] - Pika Labs stands out for its user-friendly technology, generating smooth product demonstration videos while allowing customization of backgrounds, text, and voice overlays [5] Group 3: Impact and Future Outlook - AI-generated video technology significantly enhances the production efficiency and quality of visual content in cross-border e-commerce [7] - Merchants can choose suitable tools based on their product characteristics, target markets, and resource investments for scalable video content generation [7] - Ongoing technological advancements are expected to further lower the barriers to cross-border marketing and drive the visual content upgrade in the global e-commerce market [7]

Core Points - The stock of Beijing Geling Deep Vision Technology Co., Ltd. (格灵深瞳) fell by 5.08% on September 26, closing at 16.99 CNY per share with a trading volume of 1.23 billion CNY and a turnover rate of 2.73%, resulting in a total market capitalization of 4.4 billion CNY [1] - The company, established on August 16, 2013, and listed on March 17, 2022, specializes in integrating advanced computer vision and big data analysis technologies into various applications, providing AI products and solutions for urban management, smart finance, commercial retail, sports health, and rail transit operations [1] - The revenue composition of the company includes 58.58% from AI products, 26.46% from smart terminal products, and 14.95% from technical services and others [1] Shareholder Insights - The top circulating shareholder of Geling Deep Vision is the Huabao CSI Financial Technology Theme ETF (华宝中证金融科技主题ETF), which entered the top ten shareholders in the second quarter with 1.7853 million shares, accounting for 0.95% of circulating shares, resulting in an estimated floating loss of approximately 1.6246 million CNY [2] - The Huabao CSI Financial Technology Theme ETF was established on March 4, 2021, with a current scale of 5.731 billion CNY, achieving a year-to-date return of 29.1% and ranking 1868 out of 4220 in its category, while its one-year return is 116.47%, ranking 259 out of 3824 [2] Fund Management - The fund managers of the Huabao CSI Financial Technology Theme ETF are Chen Jianhua and Cao Xucheng, with Chen having a cumulative tenure of 12 years and 282 days, managing assets totaling 15.533 billion CNY, with the best fund return during his tenure being 166.28% and the worst being -49.65% [3] - Cao has a tenure of 128 days, managing assets of 22.062 billion CNY, with the best return of 91.15% and the worst return of 7.38% during his tenure [3]

Core Viewpoint - The company, Geling Deep Vision, has experienced fluctuations in its stock performance and financial metrics, indicating potential challenges in revenue generation and profitability [1][2]. Group 1: Stock Performance - On August 27, Geling Deep Vision's stock rose by 3.59%, with a trading volume of 397 million yuan [1]. - The financing buy-in amount on the same day was 41.45 million yuan, while the financing repayment was 49.10 million yuan, resulting in a net financing buy of -7.65 million yuan [1]. - As of August 27, the total balance of margin trading was 160 million yuan, which is 4.65% of the circulating market value, indicating a relatively low financing balance compared to the past year [1]. Group 2: Financial Performance - For the first half of 2025, Geling Deep Vision reported a revenue of 42.47 million yuan, reflecting a year-on-year decrease of 17.22% [2]. - The net profit attributable to the parent company was -79.85 million yuan, showing a year-on-year decline of 2.48% [2]. - As of June 30, 2025, the number of shareholders increased by 3.78% to 15,600, while the average circulating shares per person decreased by 3.65% to 12,131 shares [2]. Group 3: Company Overview - Geling Deep Vision, established on August 16, 2013, and listed on March 17, 2022, specializes in integrating advanced computer vision and big data analysis technologies into various applications [1]. - The company's main business revenue composition includes 58.58% from AI products, 26.46% from smart terminal products, and 14.95% from technical services and others [1]. - The company is located in Haidian District, Beijing, within the Zhongguancun Science City [1].

Core Insights - Pixel Cake launched a comprehensive "online + pixel cloud space + offline" ecosystem at the "All-round Creation" themed conference in Shanghai, featuring an integrated intelligent workflow and the new Pixel Cake 8.0 version [1][3] - The integrated intelligent workflow significantly reduces the time required for photography processes from three days to just three minutes, enhancing efficiency and meeting customer demands for rapid delivery [1] - The new Pixel Cake 8.0 version introduces five innovative features, including clothing change for ID photos and intelligent prop restoration, expanding the AI creative toolbox [3] Company Overview - Pixel Cake is a subsidiary of Xiamen Xiantian Technology Co., Ltd., focusing on the intelligent application of computer vision technology in creative design, providing commercial-grade intelligent imaging solutions for commercial photography [1] - The company aims to streamline the workflow in commercial photography, allowing for a more dynamic and engaging presentation of images through various templates and effects [1] Industry Trends - According to iResearch data, as of March and April this year, AI tool users accounted for 38% of domestic AI users, indicating a growing trend in AI adoption within the commercial photography sector [5] - The commercialization of AI in photography is driven by the need to control costs and increase revenue, with applications including AI retouching, virtual scene generation, and personalized style transfer [5]

Core Insights - By 2024, China's AI-related paper citation volume is expected to account for 40.2% of the global total, rapidly catching up to the United States at 42.9% [1][8] - The report titled "Scientific Intelligence White Paper 2025" analyzes the integration of AI and scientific research across seven major research fields, covering 28 directions and nearly 90 key issues [1] - The report highlights the dual promotion and deep integration of AI innovation and scientific research, termed "AI for Science" [1] Research Trends - The number of global AI journal papers has surged nearly threefold over the past decade, from 308,900 to 954,500, with an average annual growth rate of 14% [7] - The share of core AI fields, such as algorithms and machine learning, has decreased from 44% to 38%, while the share of scientific intelligence has increased by 6 percentage points, with an annual growth rate rising from 10% before 2020 to 19% after [7] - China’s AI publication volume increased from 60,100 in 2015 to 300,400 in 2024, representing 29% of the global total [7][8] Citation Impact - The citation volume of AI-related papers in the U.S. reached 302,200 in 2020, while China's citations rose from 10,300 in 2015 to 144,800 in 2020, surpassing the EU for the first time in 2021 [8] - By 2024, China is projected to account for 41.6% of global AI citations in patents, policy documents, and clinical trials, significantly leading the field [8] Country-Specific Trends - China has a leading position in the intersection of AI with earth and environmental sciences, and has surpassed in AI with mathematics, material sciences, and humanities since 2019 [9] - The U.S. and EU maintain advantages in AI and life sciences, with China ranking third in this area [9] - India shows significant progress across all fields, currently ranking third in earth and environmental sciences, engineering, and humanities [9]