Image Processing Engineer, HDR & Tone Mapping

About the company

Efference provides high-performance perception and compute infrastructure for training and deploying robot models in the cloud. Our vision is to make robotic deployments scalable with shared compute.

We approach the problem by owning the entire path from the image sensor to the datacenter GPU. We shrink the load at the sensor, fix the uplink so robot traffic is fast and reliable, and clock synchronize the system so observations arrive time-aligned, ready for inference.

About the role

You'll develop HDR video processing that preserves useful scene information as robots move through changing light. You'll work on exposure fusion and tone mapping with particular attention to temporal stability, motion artifacts, and the effect of image processing on compression and robot perception.

What you’ll do

• Develop global, local, and adaptive tone-mapping algorithms for continuous video, balancing shadow detail, highlight retention, contrast, and temporal consistency.
• Work with sensor and ISP engineers on HDR capture and exposure fusion, accounting for readout timing, motion, saturation, and noise.
• Diagnose and reduce flicker, brightness pumping, halos, ghosting, and unstable local contrast during movement and abrupt lighting transitions.
• Build test sequences covering backlighting, reflective surfaces, low light, LED flicker, and transitions between indoor and outdoor illumination.
• Measure how tone curves and fusion choices affect codec bitrate, feature tracking, depth, and policy performance with compression and robotics engineers.
• Prototype algorithms in Python and deploy efficient embedded implementations, maintaining temporal regression tests and traceable parameter changes.

What we are looking for

• Experience developing HDR imaging, tone mapping, exposure fusion, or related camera algorithms for video.
• Strong understanding of image formation, sensor dynamic range, exposure control, color processing, and noise.
• Practical knowledge of temporal filtering and motion handling, including why a method that works on one image can fail on a sequence.
• Strong C++ and Python skills and experience profiling algorithms against memory and compute limits.
• Ability to design repeatable image-quality experiments and connect observed artifacts to their causes in the processing pipeline.

Nice to Have

• Experience with multi-exposure or sensor-level HDR, RAW/Bayer processing, and ISP hardware blocks.
• Work on learned tone mapping, temporal neural models, or quantized deployment on embedded accelerators.
• Knowledge of HDR transfer functions and formats, video compression, or the sensitivity of vision models to photometric changes.