We lifted 10 functions out of this paper's own repositories and ran 10 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 |
|---|---|---|
| richardsunnymeng/latentwatermark | canonical | 10 of 10 |
| Function | Status | Where it lives |
|---|---|---|
| collate_fn | Ran | richardsunnymeng/latentwatermark/utils/collate_fn.py pointer only (licence: NONE) · get_code("3907a9f2cba334bb") |
| conv1x1 | Ran | richardsunnymeng/latentwatermark/models/extractors/naive.py pointer only (licence: NONE) · get_code("2a80220dabcb742a") |
| conv3x3 | Ran | richardsunnymeng/latentwatermark/models/extractors/naive.py pointer only (licence: NONE) · get_code("d33c6bd60a77ad21") |
| get_param_groups_and_shapes | Ran | richardsunnymeng/latentwatermark/utils/fp16_util.py pointer only (licence: NONE) · get_code("1519c26f915f04ae") |
| load_and_precess_imgs | Ran | richardsunnymeng/latentwatermark/extract.py pointer only (licence: NONE) · get_code("0aa86a0369202eac") |
| make_master_params | Ran | richardsunnymeng/latentwatermark/utils/fp16_util.py pointer only (licence: NONE) · get_code("751bef182a1751e2") |
| msg_bce_loss | Ran | richardsunnymeng/latentwatermark/loss/loss_fn.py pointer only (licence: NONE) · get_code("475678a8f4905fe7") |
| msg_lse_loss | Ran | richardsunnymeng/latentwatermark/loss/loss_fn.py pointer only (licence: NONE) · get_code("82d651ea5f8175c5") |
| msg_reg_loss | Ran | richardsunnymeng/latentwatermark/loss/loss_fn.py pointer only (licence: NONE) · get_code("fa1f406db7aa603e") |
| unflatten_master_params | Ran | richardsunnymeng/latentwatermark/utils/fp16_util.py pointer only (licence: NONE) · get_code("64fff1e30802b815") |
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Watermarking is a tool for actively identifying and attributing the images generated by latent diffusion models. Existing methods face the dilemma of image quality and watermark robustness. Watermarks with superior image quality usually have inferior robustness against attacks such as blurring and JPEG compression, while watermarks with superior robustness usually significantly damage image quality. This dilemma stems from the traditional paradigm where watermarks are injected and detected in pixel space, relying on pixel perturbation for watermark detection and resilience against attacks. In this paper, we highlight that an effective solution to the problem is to both inject and detect watermarks in the latent diffusion space, and propose Latent Watermark with a progressive training strategy. It weakens the direct connection between quality and robustness and thus alleviates their contradiction. We conduct evaluations on two datasets and against 10 watermark attacks. Six metrics measure the image quality and watermark robustness. Results show that compared to the recently proposed methods such as StableSignature, StegaStamp, RoSteALS, LaWa, TreeRing, and DiffuseTrace, LW not only surpasses them in terms of robustness but also offers superior image quality. Our code will be available at https://github.com/RichardSunnyMeng/LatentWatermark.
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2404.00230")
get_code_for_paper("2404.00230")
have("2404.00230")
Connect an agent — have() is free.