Brahma Consulting Group
Computer Vision Engineer
Brahma Consulting Group, San Francisco, California, United States, 94199
At Brahma Consulting Group, we specialize in connecting top-tier talent with innovative companies across the tech landscape. Were currently partnering with one of our clients to hire a Computer Vision Engineer with experience in Robotics and Perception.
As a Computer Vision Engineer, youll design, develop, and deploy cutting-edge object detection and tracking pipelines, optimize models for edge devices, and collaborate on robust autonomy solutions for outdoor robotics.
Responsibilities
Architect and implement pipelines for object detection, semantic/instance segmentation, and tracking in dynamic scenes Train and evaluate models using PyTorch or TensorFlow, leveraging curated, noisy, or weakly-labeled data Optimize deep learning models for edge deployment (TensorRT, pruning, quantization, inference profiling) Create pre-processing, augmentation, and post-processing methods for real-time performance Manage and assess datasets provided by labeling teams Collaborate with cloud and systems engineers to deploy and iterate models on Jetson platforms (Orin, Xavier, Nano) Validate models in real-world scenarios, analyze performance and failure cases Integrate perception outputs with autonomy logic and mission control systems Requirements
M.S./Ph.D. in Computer Vision, Robotics, AI, or equivalent industry experience 4 years experience with object detection, segmentation, and tracking for natural scenes Strong Python skills for ML development/deployment ROS2 experience (node development, CLI tools, message integration) Expertise in PyTorch or TensorFlow, model customization and transfer learning Experience deploying models on Jetson devices; optimizing for edge constraints Familiarity with TensorRT, torchscript, and related tools Experience with embedded Linux development and debugging Knowledge of AWS or GCP for data management/training Experience with versioning/model tracking tools (DVC, Weights & Biases, MLflow)
#J-18808-Ljbffr
Architect and implement pipelines for object detection, semantic/instance segmentation, and tracking in dynamic scenes Train and evaluate models using PyTorch or TensorFlow, leveraging curated, noisy, or weakly-labeled data Optimize deep learning models for edge deployment (TensorRT, pruning, quantization, inference profiling) Create pre-processing, augmentation, and post-processing methods for real-time performance Manage and assess datasets provided by labeling teams Collaborate with cloud and systems engineers to deploy and iterate models on Jetson platforms (Orin, Xavier, Nano) Validate models in real-world scenarios, analyze performance and failure cases Integrate perception outputs with autonomy logic and mission control systems Requirements
M.S./Ph.D. in Computer Vision, Robotics, AI, or equivalent industry experience 4 years experience with object detection, segmentation, and tracking for natural scenes Strong Python skills for ML development/deployment ROS2 experience (node development, CLI tools, message integration) Expertise in PyTorch or TensorFlow, model customization and transfer learning Experience deploying models on Jetson devices; optimizing for edge constraints Familiarity with TensorRT, torchscript, and related tools Experience with embedded Linux development and debugging Knowledge of AWS or GCP for data management/training Experience with versioning/model tracking tools (DVC, Weights & Biases, MLflow)
#J-18808-Ljbffr