Pith. sign in

REVIEW 5 cited by

PointCLIP: Point Cloud Understanding by CLIP

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2112.02413 v1 pith:FFK4NMUW submitted 2021-12-04 cs.CV cs.AIcs.RO

PointCLIP: Point Cloud Understanding by CLIP

classification cs.CV cs.AIcs.RO
keywords pointclipclipcloudpointfew-shotperformanceadapterknowledge
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Recently, zero-shot and few-shot learning via Contrastive Vision-Language Pre-training (CLIP) have shown inspirational performance on 2D visual recognition, which learns to match images with their corresponding texts in open-vocabulary settings. However, it remains under explored that whether CLIP, pre-trained by large-scale image-text pairs in 2D, can be generalized to 3D recognition. In this paper, we identify such a setting is feasible by proposing PointCLIP, which conducts alignment between CLIP-encoded point cloud and 3D category texts. Specifically, we encode a point cloud by projecting it into multi-view depth maps without rendering, and aggregate the view-wise zero-shot prediction to achieve knowledge transfer from 2D to 3D. On top of that, we design an inter-view adapter to better extract the global feature and adaptively fuse the few-shot knowledge learned from 3D into CLIP pre-trained in 2D. By just fine-tuning the lightweight adapter in the few-shot settings, the performance of PointCLIP could be largely improved. In addition, we observe the complementary property between PointCLIP and classical 3D-supervised networks. By simple ensembling, PointCLIP boosts baseline's performance and even surpasses state-of-the-art models. Therefore, PointCLIP is a promising alternative for effective 3D point cloud understanding via CLIP under low resource cost and data regime. We conduct thorough experiments on widely-adopted ModelNet10, ModelNet40 and the challenging ScanObjectNN to demonstrate the effectiveness of PointCLIP. The code is released at https://github.com/ZrrSkywalker/PointCLIP.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. CR-Refiner: An Object-Centric Optimal Transport Reranker for Edit-Conditioned 3D Scene Retrieval

    cs.CV 2026-07 conditional novelty 6.0

    An unbalanced optimal-transport reranker with structural priors and an LLM verifier improves hard-subset 3D scene retrieval, evaluated on the new synthetic 3D-CER benchmark.

  2. From Geometric Labels to Semantic Understanding of Indoor Building Components Using Multimodal Large Language Models

    cs.CV 2026-07 conditional novelty 6.0

    Building-MLLM, with PIE, GPR, fixed prefix and multi-dimensional LoRA plus a 37k-pair synthetic dataset, reaches 88/65/68 % on recognition, captioning and multi-engineering QA for 47 indoor component categories.

  3. O$^3$Afford: One-Shot 3D Object-to-Object Affordance Grounding for Generalizable Robotic Manipulation

    cs.RO 2025-09 conditional novelty 6.0

    A one-shot training regime with DINOv2-enriched point clouds and joint cross-attention predicts 3D object-to-object affordance maps that guide optimization-based robotic manipulation.

  4. OpenM3D: Open Vocabulary Multi-view Indoor 3D Object Detection without Human Annotations

    cs.CV 2025-08 conditional novelty 6.0

    A single-stage image-based detector that, trained with pseudo boxes from SAM segments and CLIP features, detects and classifies arbitrary indoor objects in 3D at 0.3 seconds per scene.

  5. PASR: Pose-Aware 3D Shape Retrieval from Occluded Single Views

    cs.CV 2026-04 unverdicted novelty 5.0

    PASR performs pose-aware analysis-by-synthesis by aligning 3D projections with DINOv3 patch features, outperforming prior methods on clean and occluded retrieval while also handling pose estimation and classification.