Add TrustMark Q model files

LICENSE

@@ -0,0 +1,25 @@
+Copyright 2023 Adobe
+All Rights Reserved.
+
+NOTICE: Adobe permits you to use, modify, and distribute this file in
+accordance with the terms of the license agreement accompanying it.
+
+MIT License
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -0,0 +1,33 @@
+---
+license: other
+library_name: trustmark
+tags:
+- watermarking
+- image-watermarking
+- invisible-watermark
+- trustmark
+- provenance
+---
+
+# TrustMark Q model files for watermark toolkit
+
+This repository mirrors the TrustMark `Q` model files used by the local watermark toolkit integration.
+
+Files:
+
+- `encoder_Q.ckpt`
+- `decoder_Q.ckpt`
+- `trustmark_Q.yaml`
+
+The model is used through the Python package `trustmark>=0.9.1` and the toolkit's `TrustMarkEngine`.
+
+Original project: https://github.com/adobe/trustmark
+
+Industrial pattern used by this toolkit:
+
+```text
+payload -> SQLite/backend registry -> 8-character watermark_id -> TrustMark image watermark
+image -> TrustMark decode watermark_id -> registry lookup -> full payload
+```
+
+The image carries only the short `watermark_id`; sensitive business payloads should remain in your registry/database.

decoder_Q.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e3d9cea5406a26590735719f8f15cb10802b11852ae69047eaf4cf17214df781
+size 47652460

encoder_Q.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dc382c3f6b4fd568b27d6fbb763d6ffc2d2587126d84afe9d5ee95b4c5d99826
+size 17302074

manifest.json

@@ -0,0 +1,11 @@
+{
+  "source_package": "trustmark==0.9.1",
+  "model_type": "Q",
+  "files": [
+    "decoder_Q.ckpt",
+    "encoder_Q.ckpt",
+    "LICENSE",
+    "README.md",
+    "trustmark_Q.yaml"
+  ]
+}

trustmark_Q.yaml

@@ -0,0 +1,85 @@
+model:
+  target: trustmark.model.TrustMark_Arch
+  params:
+    cover_key: "image"
+    secret_key: "secret"
+    secret_len: 100
+    resolution: 256
+    use_ema: False
+    lr_scheduler: CosineAnnealingRestartCyclicLR
+
+    secret_encoder_config:
+      target: trustmark.unet.Unet1
+      params:
+        width: 32
+        ndown: 4
+        nmiddle: 1
+        activ: silu
+
+    secret_decoder_config:
+      target: trustmark.unet.SecretDecoder
+      params:
+        arch: resnet50
+
+    discriminator_config:
+      target: trustmark.munit.MsDCDisGP
+      params:
+        gp_weight: 10.0
+        num_scales: 1
+        norm: none
+
+    loss_config:
+      target: trustmark.loss.ImageSecretLoss
+      params:
+        recon_type: ffl+yuv
+        recon_weight: 1.5
+        perceptual_weight: 1.0
+        kl_weight: 0.0
+        secret_weight: 20.0
+        generator_weight: 0.5
+        discriminator_weight: 1.0
+        generator_update_freq: 2
+        max_image_weight_ratio: 27.5
+
+    noise_config:
+      target: trustmark.utils.transformations2.TransformNet
+      params:
+        ramp: 10000
+        severity: high
+        crop_mode: resized_crop
+        gamma: false  # this cause issue
+
+data:
+  target: trustmark.utils.imgcap_dataset.DataModuleFromConfig
+  params:
+    batch_size: 32
+    num_workers: 4
+    wrap: false
+    use_worker_init_fn: true
+
+lightning:
+  callbacks:
+    image_logger:
+      target: trustmark.logger.ImageLogger
+      params:
+        batch_frequency: 5000
+        max_images: 4
+        increase_log_steps: False
+        fixed_input: True  # log the same image batch
+    progress_bar:
+      target: lightning.pytorch.callbacks.TQDMProgressBar
+      params:
+        refresh_rate: 100
+    checkpoint:
+      target: lightning.pytorch.callbacks.ModelCheckpoint
+      params:
+        verbose: true
+        filename: '{epoch:06}-{step:09}'
+        every_n_train_steps: 10000
+
+  trainer:
+    benchmark: True
+    base_learning_rate: 4e-6
+    lr_mult: true
+    accumulate_grad_batches: 1
+    max_epochs: 150