SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 1611.01704 · 2016

End-to-end Optimized Image Compression

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 20 functions out of this paper's own repositories and ran 9 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.

RepositoryRoleRan
liujiaheng/iclr_17_compression pwc_unofficial 5 of 12
copy not recorded — 4 of 4
treammm/Compression pwc_unofficial 0 of 4
FunctionStatusWhere it lives
Average Ran this paper's copy was not recorded; identical code first harvested from InterDigitalInc/CompressAI
pointer only · get_code("390aec91fdddfcac")
BoolConvert Ran this paper's copy was not recorded; identical code first harvested from InterDigitalInc/CompressAI
pointer only · get_code("d729e65bfe314ae3")
CalcuPSNR Ran liujiaheng/iclr_17_compression/models/basics.py
code served (permissive licence) · get_code("c9947c33e3a54f62")
get_loader Ran liujiaheng/iclr_17_compression/datasets.py
code served (permissive licence) · get_code("330dd4378c3e3a35")
get_train_loader Ran liujiaheng/iclr_17_compression/datasets.py
code served (permissive licence) · get_code("4e3ec336e202299e")
inverse_dict Ran this paper's copy was not recorded; identical code first harvested from InterDigitalInc/CompressAI
pointer only · get_code("5dff248eb7143d81")
print_dict Ran this paper's copy was not recorded; identical code first harvested from gergely-flamich/relative-entropy-coding
pointer only · get_code("73b28d07dc4d8995")
relu Ran liujiaheng/iclr_17_compression/models/basics.py
code served (permissive licence) · get_code("ba24f201a717509b")
ssim Ran liujiaheng/iclr_17_compression/models/ms_ssim_torch.py
code served (permissive licence) · get_code("d9e269e86cc2b394")
MultiScaleSSIM Not yet run liujiaheng/iclr_17_compression/metric.py
code served (permissive licence) · get_code("a5a05f863cc1b904")
gaussian_filter Not yet run liujiaheng/iclr_17_compression/models/ms_ssim_torch.py
code served (permissive licence) · get_code("569f86999a28550f")
irdft_matrix Not yet run treammm/Compression/tensorflow_compression/python/ops/spectral_ops.py
code served (permissive licence) · get_code("e1be96227f48724d")
load_model Not yet run liujiaheng/iclr_17_compression/model.py
code served (permissive licence) · get_code("67b83d1f47e01c3f")
lower_bound Not yet run treammm/Compression/tensorflow_compression/python/ops/math_ops.py
code served (permissive licence) · get_code("a863fa9d143a5fb3")
ms_ssim Not yet run liujiaheng/iclr_17_compression/models/ms_ssim_torch.py
code served (permissive licence) · get_code("7ea107a5705e6f95")
msssim Not yet run liujiaheng/iclr_17_compression/metric.py
code served (permissive licence) · get_code("c46a2076f53009e8")
psnr Not yet run liujiaheng/iclr_17_compression/metric.py
code served (permissive licence) · get_code("94c8354fe1b4c307")
same_padding_for_kernel Not yet run treammm/Compression/tensorflow_compression/python/ops/padding_ops.py
code served (permissive licence) · get_code("680ff131d22fbc0b")
upper_bound Not yet run treammm/Compression/tensorflow_compression/python/ops/math_ops.py
code served (permissive licence) · get_code("fc9f7ea55d17f349")
yuv_import_444 Not yet run liujiaheng/iclr_17_compression/models/basics.py
code served (permissive licence) · get_code("2784ebaef09c1269")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

We describe an image compression method, consisting of a nonlinear analysis transformation, a uniform quantizer, and a nonlinear synthesis transformation. The transforms are constructed in three successive stages of convolutional linear filters and nonlinear activation functions. Unlike most convolutional neural networks, the joint nonlinearity is chosen to implement a form of local gain control, inspired by those used to model biological neurons. Using a variant of stochastic gradient descent, we jointly optimize the entire model for rate-distortion performance over a database of training images, introducing a continuous proxy for the discontinuous loss function arising from the quantizer. Under certain conditions, the relaxed loss function may be interpreted as the log likelihood of a generative model, as implemented by a variational autoencoder. Unlike these models, however, the compression model must operate at any given point along the rate-distortion curve, as specified by a trade-off parameter. Across an independent set of test images, we find that the optimized method generally exhibits better rate-distortion performance than the standard JPEG and JPEG 2000 compression methods. More importantly, we observe a dramatic improvement in visual quality for all images at all bit rates, which is supported by objective quality estimates using MS-SSIM.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("1611.01704")
get_code_for_paper("1611.01704")
have("1611.01704")

Connect an agent — have() is free.