GO-PRE: Goal-Oriented Next-Best-View Selection via Predictive Rendering Entropy for Active 3D Reconstruction

Active 3D reconstruction relies on active view selection to maximize reconstruction fidelity under limited capture budgets. However, most existing methods rely on surrogate signals—such as parameter uncertainty or geometric heuristics—which are often misaligned with the ultimate goal: the fidelity of rendered predictions. We propose GO-PRE, a goal-oriented next-best-view selection framework that explicitly targets information gain in the prediction space. Specifically, we formulate the objective as maximizing the reduction of the average marginal predictive entropy over a user-specified target view manifold. GO-PRE supports interactive goal specification and yields an efficient acquisition rule that enables real-time computation of information gain. Extensive experiments across benchmarks demonstrate that GO-PRE consistently improves active reconstruction performance and provides more reliable uncertainty quantification compared to state-of-the-art methods.

To build a 3D model of a scene, a camera or robot must photograph it from many angles — but capturing every possible angle is slow and costly, so it must choose a small number of viewpoints wisely. Most existing methods pick these viewpoints using indirect clues, such as how uncertain the model's internal numbers are. The problem is that these clues don't always match what we actually care about: how good the final rendered pictures look. We developed GO-PRE, a method that chooses each new viewpoint by directly predicting which one will most improve the clarity of the final images. It also lets a user point to a specific region they care about, so the system focuses its limited budget there instead of treating the whole scene equally. Across several standard benchmarks, GO-PRE produced sharper reconstructions than leading methods and gave more trustworthy estimates of where a reconstruction is still unreliable. This could help autonomous systems — such as inspection drones or search-and-rescue robots — reconstruct environments faster and more reliably.