Yan Wang, Jun Zhang, Tongda Xu, Dailan He, Hongwei Qin, Xinjie Zhang, Xingtong Ge, Guo Lu, Jing Geng
We lifted 7 functions out of this paper's own repositories and ran 7 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.
| Repository | Role | Ran |
|---|---|---|
| Xinjie-Q/GaussianImage | canonical | 7 of 7 |
| Function | Status | Where it lives |
|---|---|---|
| grad_scale | Ran | Xinjie-Q/GaussianImage/quantize.py code served (permissive licence) · get_code("a291c2776caf3a35") |
| image_path_to_tensor | Ran | Xinjie-Q/GaussianImage/train_quantize.py code served (permissive licence) · get_code("f5b2d15797cbfb2c") |
| parse_args | Ran | Xinjie-Q/GaussianImage/train_quantize.py code served (permissive licence) · get_code("98c39765afc54fc8") |
| random_quat_tensor | Ran | Xinjie-Q/GaussianImage/gaussiansplatting_3d.py code served (permissive licence) · get_code("63025c93e75efbe9") |
| ste | Ran | Xinjie-Q/GaussianImage/quantize.py code served (permissive licence) · get_code("13a834379c89aaae") |
| strip_lowerdiag | Ran | Xinjie-Q/GaussianImage/utils.py code served (permissive licence) · get_code("e3cb74fbb6292ab5") |
| strip_symmetric | Ran | Xinjie-Q/GaussianImage/utils.py code served (permissive licence) · get_code("11f871291d5ed58c") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Implicit neural representations (INRs) recently achieved great success in image representation and compression, offering high visual quality and fast rendering speeds with 10-1000 FPS, assuming sufficient GPU resources are available. However, this requirement often hinders their use on low-end devices with limited memory. In response, we propose a groundbreaking paradigm of image representation and compression by 2D Gaussian Splatting, named GaussianImage. We first introduce 2D Gaussian to represent the image, where each Gaussian has 8 parameters including position, covariance and color. Subsequently, we unveil a novel rendering algorithm based on accumulated summation. Remarkably, our method with a minimum of 3× lower GPU memory usage and 5× faster fitting time not only rivals INRs (e.g., WIRE, I-NGP) in representation performance, but also delivers a faster rendering speed of 1500-2000 FPS regardless of parameter size. Furthermore, we integrate existing vector quantization technique to build an image codec. Experimental results demonstrate that our codec attains rate-distortion performance comparable to compression-based INRs such as COIN and COIN++, while facilitating decoding speeds of approximately 2000 FPS. Additionally, preliminary proof of concept shows that our codec surpasses COIN and COIN++ in performance when using partial bits-back coding. Code is available at https://github.com/Xinjie-Q/GaussianImage.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2403.08551")
get_code_for_paper("2403.08551")
have("2403.08551")
Connect an agent — have() is free.