绣春刀
深度学习可识别显微照片中的细菌_我的网站

一 | (ECNS) -- Robots are moving into everyday service roles across Beijing in 2026, brewing coffee in under 30 seconds, cooking dozens of Chinese dishes and going on sale in shopping malls as the city's robotics industry expands. In the first half of 2026, key humanoid robot manufacturers in Beijing produced nearly 6,000 units, with output up nearly tenfold year on year. The city has more than 30 humanoid robot manufacturers, accounting for over 20% of the national total. At a 7Fresh Coffee smart unmanned store in Beijing’s Galaxy SOHO, robots handle grinding, extraction, mixing and dispensing, with each drink made in under 30 seconds. On its opening day Sunday, the store successfully completed a Guinness World Records challenge by producing 202 cups in one hour. At a robot restaurant in Beijing’s Yizhuang, an area known for driverless cars and humanoid robots,, nearly 20 robots handle tasks from greeting guests to post-meal cleaning. A robot chef can prepare about 50 classic Chinese dishes, including braised pork and stir-fried shredded potatoes, with each seasoning measured to within 0.1 gram. Consumers can also buy robots in Beijing’s shopping districts. Unitree Robotics opened its first directly operated store in Wangfujing in late April, where visitors can interact with products including the G1 humanoid robot and Go2 robot dog. (By Intern Yang Hongran, Zhang Dongfang)
。 此外,由于不同细菌的光学特性存在差异,Omnipose在克服识别问题方面表现出色。由于抗生素治疗或细菌间侵袭过程中产生的化学物质的拮抗作用,该软件不会轻易被细胞形状的极端变化所欺骗。事实上,该程序表明它甚至可在大肠杆菌试验中检测细胞中毒。大多数细菌是球形或棒状,但有些具有其他基本形式,例如扭曲的螺旋形。Omnipose还可识别更精细的细菌,这些细菌具有细长的形状,其物理特征会使深度学习工具难以判断图像中存在哪些细菌。

二 | 在分析来自杂草拟南芥根的细胞时,Omnipose在这个3D样本中确实显示出一些优势。

三 | 为了探究它是否也可成为其他依赖显微镜的生物甚至非生命科学领域的多功能工具,研究人员在超小型蛔虫——秀丽隐杆线虫的显微照片上试用了该程序,秀丽隐杆线虫是遗传、神经科学、发育和微生物行为研究中的重要生物。像一些细菌一样,这种生物具有细长的形状,它也可像许多其它蠕虫一样扭曲自己。结果表明,无论它的各种伸展、收缩和其他运动如何,Omnipose都可挑选出秀丽隐杆线虫。

四 | 这种能力在延时跟踪秀丽隐杆线虫运动期间的神经研究中可能是有用的。在设计像Omnipose这样的工具时,研究人员正在研究单像素精度的尺度来定义细胞的边界。

五 | 这是因为大多数细菌细胞体图像仅由少量像素组成。研究人员解释说,在图像中定义边界称为分割。他们通过深度神经网络、高精度分割算法开发了Ominpose。实验表明Omnipose具有前所未有的分割精度。

六 | 研究人员称,Omnipose在各种细胞形态和模式上的高性能,可能会从以前无法访问的显微镜图像中解开信息,这或将改变生物图像分析的游戏规则。
Current article:http://3sea.jiaopeikongbenlo.pics/yqeq0/vip.html
Published on:00:00:00
















