Perspective Plane Program Induction From a Single Image

SetCameraPose(pitch = -42, yaw = -15) for i in range(0, 4): for j in range(0, 4): if i + j >= 3: draw( , x = 18.7 * i + 9.7 * j, y = 19.7 * j) (a) Input Image (d) Failed Regularity Inference (e) Failed Perspective Correction PlaneNet (Learning-Based) RPD (Non-Perspective-Aware) (c) Perspective Plane Program Figure 1: Perspective effects and scene structure regularity are ubiquitous in natural images (a). To detect such regularity, one may directly apply regularity structure detection (RPD) [21] to natural images, but this often fails due to the existence of perspective effects (d). Attempting to remedy this, one may perform perspective correction as an independent preprocessing step, but perspective correction often relies on line and/or vanishing point cues, and fails when such cues are missing (e). We observe that these two tasks are interconnected: image regularity serves as a new perspective correction cue, and regularity detection, in turn, also benefits from perspective correction. Thus, we propose to jointly solve perspective correction and regularity structure detection (b) by simultaneously seeking the program and perspective parameters that best describe the image (c).