摘要: Computer vision professionals develop systems that monitor endangered fish species, alert car dealerships of potential theft, and track inventory on retailers' shelves, to name a few examples. While their products vary, their day-... 展开 Computer vision professionals develop systems that monitor endangered fish species, alert car dealerships of potential theft, and track inventory on retailers' shelves, to name a few examples. While their products vary, their day-to-day work practices rarely do. Most work in small, co-located, multi-disciplinary teams and rapidly iterate systems built with algorithms, products, and services provided by similar teams but at different companies. Their examples challenge global software engineering research to look beyond the social and technical coordination work of large, internal software development teams. Their stories, culled from ethnographic interviews with eighty computer vision engineers and research scientists, echo the heady days of 1980s Silicon Valley when Bay Area social networks and the global migration of professional talent fueled the rapid growth of the high technology industry. These engineers, then and now, migrate from task to task, versus the task from engineer to engineer. 收起