3 ms·

We rendered the one million part ABC dataset from Deep Geometry, and open-sourced the data. We also built a fun demo with the following pipeline: CAD > render >

We rendered the one million part ABC dataset from Deep Geometry, and open-sourced the data. We also built a fun demo with the following pipeline: CAD > render > caption > embed. Open-sourced dataset: https://huggingface.co/datasets/daveferbear/3d-model-images-embeddings https://huggingface.co/datasets/daveferbear/3d-model-images-... Blog writeup: https://www.finalrev.com/blog/embedding-one-million-3d-models https://www.finalrev.com/blog/embedding-one-million-3d-model...

MichaelZuo 8mo ago
The search function doesn’t seem to work at all, it provides nonsensical results. For example if I search “supercolumns” I get regular household furniture.
- hacker_homie 8mo ago
  Yeah I think the embedding are describing what can be seen from a picture of the model not what it is or what it is used for. search some things work like "Fan" but others don't so you can search for "plate with 5 holes" but not for specific engine part cover.
wlesieutre 8mo ago
When I search for "chair" I get 48 results, about 3 of which are actually chair-like. - Some are clearly miscategorized - ABC-00131096 is a coffee table but has a very detailed description of its chair attributes. - Many others are weird nonsense geometry, like ABC-00991744, ABC-00807798, ABC-00349255 or ABC-00822766. - Some have a partially accurate description (if you pretend it's a chair), like ABC-00685912 has a blocky geometric structure with a horizontal piece off a vertical piece, but then it starts talking about an armrest on one side that doesn't exist at all. - ABC-00388826 is a silhouette of a cat, which the description misses completely, and I don't see how you would sit on this "unique chair design characterized by its fluid and sculptural form." Overall the descriptions are pretty useless and ascribe a lot of chairness to things that are not chairs. Is a dataset with this much junk in it good for something?

MichaelZuo 8mo ago

The search function doesn’t seem to work at all, it provides nonsensical results. For example if I search “supercolumns” I get regular household furniture.

hacker_homie 8mo ago
Yeah I think the embedding are describing what can be seen from a picture of the model not what it is or what it is used for. search some things work like "Fan" but others don't so you can search for "plate with 5 holes" but not for specific engine part cover.

wlesieutre 8mo ago

When I search for "chair" I get 48 results, about 3 of which are actually chair-like. - Some are clearly miscategorized - ABC-00131096 is a coffee table but has a very detailed description of its chair attributes. - Many others are weird nonsense geometry, like ABC-00991744, ABC-00807798, ABC-00349255 or ABC-00822766. - Some have a partially accurate description (if you pretend it's a chair), like ABC-00685912 has a blocky geometric structure with a horizontal piece off a vertical piece, but then it starts talking about an armrest on one side that doesn't exist at all. - ABC-00388826 is a silhouette of a cat, which the description misses completely, and I don't see how you would sit on this "unique chair design characterized by its fluid and sculptural form." Overall the descriptions are pretty useless and ascribe a lot of chairness to things that are not chairs. Is a dataset with this much junk in it good for something?