PointNet
Architectures AIMech:PointNet DRAFT
A neural architecture for unordered point clouds that applies shared pointwise feature extraction and symmetric max pooling to produce a global representation, with a per-point prediction extension for segmentation.
Identity Notes & Concept Scope
This is the original PointNet architecture, including its classification and segmentation paths. It is distinct from PointNet++'s hierarchical local grouping and from a generic set-pooling model.
Claims & Grounded Evidence
1 Claim
Mechanism
#architecture-mechanism
Shared multilayer perceptrons encode points; max pooling aggregates their features. The segmentation path concatenates the global descriptor with point features before shared prediction layers. Learned input and feature transformations provide alignment modules.
⚠️ Curated Limitations & Negative Results
- Permutation invariance of a global output and permutation equivariance of per-point predictions are different properties.
- Learned alignment modules do not establish exact invariance to every geometric transformation; no universal accuracy or robustness claim is curated.
Artificial Intelligence Ontology (AIO) Grounding
Mapped AIO Classes
| AIO Target Class | Mapping Relation | Mapping Rationale |
|---|---|---|
AIO:ArtificialNeuralNetwork
Artificial Neural Network
|
broad_match | Shared neural feature extractors, learned transforms and neural prediction paths make this a specific artificial neural architecture. |
Evidence Sources & Primary Literature
1 Source| Source ID | Publication / Document | Type / Access | Reference & Checksum |
|---|---|---|---|
original-paper
|
Charles R. Qi, Hao Su, Kaichun Mo, Leonidas J. Guibas
|
primary paper
full_text
|
sha256:7c4c69310de5…
|
Curation Provenance
Curator: Codex
Method: ai_assisted
Model: GPT-6
Created: 2026-10-11T04:19:31Z
Updated: 2026-10-11T04:19:31Z
Source File:
data/curated/PointNet.yaml