Almost immediately after our live podcast with Ruilong Li, he was back on GitHub, pushing through pull requests and releasing gsplat 1.3.
This update includes several smaller improvements, but notably, it officially introduces the compression method into the repository. What stands out to me in this release is the inclusion of orthographic projection.
Here is the full changelog:
docs: update README.md by @eltociear in #322
Consolidate simple_trainer.py and simple_trainer_mcmc.py by @brentyi in #325
Compressing gaussian splats by @jefequien in #309
Simplify
tyro
usage insimple_trainer.py
by @brentyi in #332Cleanup C/CUDA code for C users by @liruilong940607 in #333
Patch to make Strategy compatible with splatfacto by @kerrj in #334
Fix minor typo by @zhiguotang5 in #342
correct distCoeffs of SIMPLE_RADIAL by @Master-cai in #337
Add np.random.choice fallback for many Gaussians exceeding torch.multinomial limits by @soskek in #338
data parser for t&t dataset by @liruilong940607 in #347
Added orthographic projection by @VladislavZavadskyy in #349
Fix bug related to refine_scale2d and add pause_refine_after_reset to default strategy by @jb-ye in #354
tt compression script by @liruilong940607 in #357
Update to 1.3.0 by @liruilong940607 in #366
gpslat continues to be Apache 2.0 licensed and welcoming of contributions! To get started, check out their Github.