The Flow has not summarised this story yet — read it at arXiv Machine Learning.
arXiv:2609.19853v2 Announce Type: replace-cross Abstract: Between a screenplay and a film sits a planning problem that is spatial first: who stands where, and what a camera sees from where it stands....
CALIPER is a new benchmark that tests whether pretrained visual encoders can infer physical properties such as mass and friction from images. The test involves striking an object twice at known speeds, showing a third strike only up to contact, and asking a linear readout on frozen features to predict how far the object slides. Results show that in clean, fixed‑camera scenes all representations perform similarly, but when camera, lighting, and clutter are varied, only encoders that truly infer physics—like V‑JEPA 2—maintain performance, while random or raw pixel representations fail.
PACE (Precise AI Cinematic Expression) is a typed specification that captures a film’s spatial plan—screenplay evidence, characters, props, locations, and camera actions—at script, scene, shot, or panel levels, with inheritance to lower levels. A compiler transforms this plan into prompts for diffusion models and a metrically accurate 3D scene, while a camera solver ensures the declared framing is achieved. Evaluation on the Automatic Drive screenplay and 204 external shots shows high positional accuracy and improved action capture when poses are declared, though transitions and motion remain future work.
The paper introduces a counterfactual benchmark to evaluate whether vision models understand physical constraints or merely learn rendering shortcuts. Using Mitsuba 3, it generates 5,500 synthetic images across 500 scene families, each with a clean image and ten violations (shadows, reflections, support, surface response, occlusion). Diagnostic baselines (LLaVA‑1.5‑7B, Qwen2.5‑VL‑7B, InternVL3.5‑8B) achieve 64.0–67.8 % macro‑F1 on a held‑out synthetic test set, but performance drops to 40.8 % for LLaVA‑1.5‑7B under an intervention shift, indicating reliance on rendering cues rather than true physical reasoning.
The paper challenges the common assumption that figurative paintings represent a single, recoverable 3D scene. It introduces the concept of multi-solutionness, highlighting how unobserved content and ambiguous visual cues allow multiple plausible 3D configurations. The authors present a workflow that generates multiple camera-orbit video sequences from a single painting, reconstructs each with 3D Gaussian Splatting, and fabricates the resulting point-based Gaussian scenes as physical artifacts via DreamPrinting, thereby making the non-uniqueness of interpretations explicit and inspectable.
arXiv:2609.17450v1 Announce Type: new Abstract: Novel-view synthesis from a single image is a fundamentally ambiguous problem. As the camera moves away from the input viewpoint, previously hidden reg...