Computer Vision jobs

Computer vision hiring has split in two: the classic detection-and-segmentation work that runs factories, farms, and radiology labs, and the newer multimodal wave where vision is one input to a language model. Teams still ship YOLO variants and ViTs on edge hardware while their colleagues wire images into LLM context windows. Both ends of that split are represented here.

Expect PyTorch as table stakes, with OpenCV, CUDA, and TensorRT close behind for anything real-time. Domain niches — medical imaging, autonomous driving, satellite imagery — pay a premium for candidates who know the sensor as well as the model. Remote work is common except where the hardware lives, and postings usually say which kind they are.