From 127223b11401125a9fce5ce269bb0546ee4de6e8 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ruilong=20Li=28=E6=9D=8E=E7=91=9E=E9=BE=99=29?= Date: Sun, 9 Oct 2022 09:27:43 -0700 Subject: [PATCH] sync pref (#60) --- README.md | 2 +- docs/source/examples/dnerf.rst | 7 ++++--- docs/source/examples/ngp.rst | 2 +- docs/source/examples/unbounded.rst | 2 +- docs/source/examples/vanilla.rst | 3 ++- docs/source/index.rst | 2 +- 6 files changed, 10 insertions(+), 8 deletions(-) diff --git a/README.md b/README.md index 886544b2..be4c2c1f 100644 --- a/README.md +++ b/README.md @@ -14,7 +14,7 @@ Using NerfAcc, - The `Instant-NGP NeRF` model can be trained to *better quality* (+~0.7 PSNR) with *9/10th* of the training time (4.5 minutes) comparing to the official pure-CUDA implementation. - The `D-NeRF` model for *dynamic* objects can also be trained in *1 hour* - rather than *2 days* as in the paper, and with *better quality* (+~0.5 PSNR). + rather than *2 days* as in the paper, and with *better quality* (+~2.0 PSNR). - Both *bounded* and *unbounded* scenes are supported. **And it is pure Python interface with flexible APIs!** diff --git a/docs/source/examples/dnerf.rst b/docs/source/examples/dnerf.rst index 1f711685..56c407c9 100644 --- a/docs/source/examples/dnerf.rst +++ b/docs/source/examples/dnerf.rst @@ -5,6 +5,7 @@ See code `examples/train_mlp_dnerf.py` at our `github repository`_ for details. Benchmarks ------------ +*updated on 2022-10-08* Here we trained a 8-layer-MLP for the radiance field and a 4-layer-MLP for the warping field, (similar to the T-Nerf model in the `D-Nerf`_ paper) on the `D-Nerf dataset`_. We used train @@ -24,12 +25,12 @@ single NVIDIA TITAN RTX GPU. The training memory footprint is about 11GB. +======================+==========+=========+=======+=========+=======+========+=========+=======+=======+ | D-Nerf (~ days) | 38.93 | 25.02 | 29.25 | 32.80 | 21.64 | 31.29 | 32.79 | 31.75 | 30.43 | +----------------------+----------+---------+-------+---------+-------+--------+---------+-------+-------+ -| Ours (~ 50min) | 39.60 | 22.41 | 30.64 | 29.79 | 24.75 | 35.20 | 34.50 | 31.83 | 31.09 | +| Ours (~ 1 hr) | 39.49 | 25.58 | 31.86 | 32.73 | 24.32 | 35.55 | 35.90 | 32.33 | 32.22 | +----------------------+----------+---------+-------+---------+-------+--------+---------+-------+-------+ -| Ours (Training time)| 45min | 49min | 51min | 46min | 53min | 57min | 49min | 46min | 50min | +| Ours (Training time)| 37min | 52min | 69min | 64min | 44min | 79min | 79min | 39min | 58min | +----------------------+----------+---------+-------+---------+-------+--------+---------+-------+-------+ .. _`D-Nerf`: https://arxiv.org/abs/2011.13961 .. _`D-Nerf dataset`: https://www.dropbox.com/s/0bf6fl0ye2vz3vr/data.zip?dl=0 -.. _`github repository`: https://github.com/KAIR-BAIR/nerfacc/ +.. _`github repository`: https://github.com/KAIR-BAIR/nerfacc/tree/5637cc9a1565b2685c02250eb1ee1c53d3b07af1 diff --git a/docs/source/examples/ngp.rst b/docs/source/examples/ngp.rst index cea599af..4c45b9c4 100644 --- a/docs/source/examples/ngp.rst +++ b/docs/source/examples/ngp.rst @@ -31,5 +31,5 @@ memory footprint is about 3GB. +----------------------+-------+-------+---------+-------+-------+-------+-------+-------+-------+ .. _`Instant-NGP Nerf`: https://arxiv.org/abs/2201.05989 -.. _`github repository`: https://github.com/KAIR-BAIR/nerfacc/ +.. _`github repository`: https://github.com/KAIR-BAIR/nerfacc/tree/5637cc9a1565b2685c02250eb1ee1c53d3b07af1 .. _`Nerf-Synthetic dataset`: https://drive.google.com/drive/folders/1JDdLGDruGNXWnM1eqY1FNL9PlStjaKWi diff --git a/docs/source/examples/unbounded.rst b/docs/source/examples/unbounded.rst index 78e706e3..eea315ca 100644 --- a/docs/source/examples/unbounded.rst +++ b/docs/source/examples/unbounded.rst @@ -40,4 +40,4 @@ that takes from `MipNerf360`_. .. _`Instant-NGP Nerf`: https://arxiv.org/abs/2201.05989 .. _`MipNerf360`: https://arxiv.org/abs/2111.12077 .. _`Nerf++`: https://arxiv.org/abs/2010.07492 -.. _`github repository`: https://github.com/KAIR-BAIR/nerfacc/ +.. _`github repository`: https://github.com/KAIR-BAIR/nerfacc/tree/5637cc9a1565b2685c02250eb1ee1c53d3b07af1 diff --git a/docs/source/examples/vanilla.rst b/docs/source/examples/vanilla.rst index a0412549..a4ffb9cb 100644 --- a/docs/source/examples/vanilla.rst +++ b/docs/source/examples/vanilla.rst @@ -5,6 +5,7 @@ See code `examples/train_mlp_nerf.py` at our `github repository`_ for details. Benchmarks ------------ +*updated on 2022-10-08* Here we trained a 8-layer-MLP for the radiance field as in the `vanilla Nerf`_. We used the train split for training and test split for evaluation as in the Nerf paper. Our experiments are @@ -28,5 +29,5 @@ conducted on a single NVIDIA TITAN RTX GPU. The training memory footprint is abo | Ours (Training time)| 58min | 53min | 46min | 62min | 56min | 42min | 52min | 49min | 52min | +----------------------+-------+-------+---------+-------+-------+-------+-------+-------+-------+ -.. _`github repository`: : https://github.com/KAIR-BAIR/nerfacc/ +.. _`github repository`: : https://github.com/KAIR-BAIR/nerfacc/tree/5637cc9a1565b2685c02250eb1ee1c53d3b07af1 .. _`vanilla Nerf`: https://arxiv.org/abs/2003.08934 diff --git a/docs/source/index.rst b/docs/source/index.rst index 01964c71..4022cc20 100644 --- a/docs/source/index.rst +++ b/docs/source/index.rst @@ -11,7 +11,7 @@ Using NerfAcc, - The `Instant-NGP Nerf`_ model can be trained to *better quality* (+~0.7 PSNR) with *9/10th* of \ the training time (4.5 minutes) comparing to the official pure-CUDA implementation. - The `D-Nerf`_ model for *dynamic* objects can also be trained in *1 hour* \ - rather than *2 days* as in the paper, and with *better quality* (+~0.5 PSNR). + rather than *2 days* as in the paper, and with *better quality* (+~2.0 PSNR). - Both *bounded* and *unbounded* scenes are supported. **And it is pure Python interface with flexible APIs!**