職位描述：1.圖像相關算法的研發(fā)，包括但不限于目標檢測，OCR，行為分析等；2.負責數據集的構建，標注，模型訓練和效果評估，推進流程的自動化，效率化；3.與產品和前后端配合，推進工程化部署和性能優(yōu)化。職位要求：1.211碩士以上，計算機/電子信息相關專業(yè)，良好的數學基礎特別是矩陣；2.會目標檢測，圖像分割，OCR中至少一種；3.會用pytorch，能寫C++和python，熟悉傳統(tǒng)圖像處理例如去畸變，對比度增強。加分項：會點云處理，3D檢測，cuda編程(能自己寫核函數)

上海西井科技股份有限公司（簡稱西井科技WESTWELL），致力于以人工智能激活多產業(yè)潛能，做全球智能服務業(yè)的開疆者。西井科技是目前為止中國首批基于正向自研無人駕駛重卡的公司，也是在全球范圍商業(yè)交付無人駕駛車輛及系統(tǒng)超過數百套規(guī)模的公司，場景已覆蓋海港、陸港、制造工廠等貨運物流，產品及服務已落地全球18個國家和地區(qū)，服務160余家客戶。秉持“取智于人，賦能予人From Human To Human”的企業(yè)哲學，西井科技憑借獨特的應用場景理解能力，已形成以WellOcean?（海港）、鐵路樞紐、陸港、空港、工廠等代表的大物流領域多場景智能解決方案，通過Qomolo?（新能源無人駕駛及智能商用車），用運力串聯(lián)起各生產要素的運營與流轉，并通過構建Loopo?（數字化生產要素流轉互聯(lián)平臺），從生產節(jié)點、貿易流程等環(huán)節(jié)，為終端用戶創(chuàng)造更高效并可持續(xù)的生產力價值。Shanghai Westwell Technology Co., Ltd (“Westwell”) is committed to activating the potential of various industries through artificial intelligence, and to be the industry pioneer of global intelligent services.Westwell has been the leading company to independently develop autonomous heavy trucks, and also the company to deliver hundreds of autonomous vehicles and sets of autonomous driving systems commercially worldwide. Westwell’s products and services cover most freight logistics scenarios, such as seaports, dry ports, manufacturing plants, spanning 160-plus clients in 18 countries and regions around the world. Adhering to the company’s Value Concept “From Human to Human” and through its extraordinary insights, Westwell has developed multi-scenario intelligent solutions for the bulk logistics fields, including seaports (dedicated seaport solution — WellOcean?), railway hubs, dry ports, airports, and factories, etc. With Qomolo? (our autonomous new energy commercial vehicle brand), Westwell connects operation and circulation within various production factors with transport capacity. With the launch of Loopo? (our digital interconnection platform for production factors), Westwell is capable of delivering more efficient and sustainable productivity value for end users in every sector, such as production nodes and trading flow, etc.