Cambrian-P: Pose-Grounded Video Understanding
For video understanding researchers, this work shows that incorporating camera pose as a lightweight signal significantly boosts spatial reasoning and general video QA, addressing a key limitation of current MLLMs.
The paper introduces Cambrian-P, a video multimodal LLM that uses per-frame camera pose tokens and a pose regression head, achieving 4.5-6.5% gains on spatial reasoning benchmarks like VSI-Bench and SOTA streaming pose estimation on ScanNet, while also improving general video QA.
Camera pose matters. The position and orientation of each viewpoint define a shared spatial coordinate frame that relates observations across video frames. Yet this signal is largely absent from multimodal LLMs (MLLMs) for video understanding, which process frames as isolated 2D snapshots, instead of the persistent scene humans perceive. We revisit pose as a lightweight supervisory signal and introduce Cambrian-P, a video MLLM augmented with per-frame learnable camera tokens and a pose regression head. With a carefully designed sampling scheme, the model achieves substantial gains of 4.5-6.5% on spatial reasoning benchmarks such as VSI-Bench, generalizes across eight additional spatial and general video QA benchmarks, and, as a byproduct, achieves state of the art streaming pose estimation on ScanNet. Surprisingly, training on pseudo-annotated poses from in-the-wild video further improves general video QA benchmarks, showing pose helps beyond spatial reasoning. Together, these results position camera pose as a fundamental signal for video models that reason about the physical world.