Looking at words and points with attention: a benchmark for text-to-shape coherence
(ICCV 2023 Workshop "AI for 3D Content Creation")

Dataset GPT2Shape

Train-val-test splits of GPT2Shape dataset can be found in folder gpt2shape:

• train
• val
• test

Each CSV file contains the ShapeNet model_id of every shape, the corresponding textual description, the name and id of the category and other metadata.

Installation

Evaluation metric CrossCoherence - inference

Inference

Evaluation metric CrossCoherence - training

Training

Human-validated test set HST

Citation

If you find this work useful, please cite our paper:

@InProceedings{Amaduzzi_2023_ICCV,
    author    = {Amaduzzi, Andrea and Lisanti, Giuseppe and Salti, Samuele and Di Stefano, Luigi},
    title     = {Looking at Words and Points with Attention: a Benchmark for text-to-Shape Coherence},
    booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) Workshops},
    month     = {October},
    year      = {2023},
    pages     = {2868-2877}
}