- Rating
- 3(+1)
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 3
Read review
Paper summary
This paper introduces SimpleMark, a minimalist audio watermarking framework that breaks away from paradigms relying on complex designs such as latent space embedding, attack simulation, and perceptual loss. By employing only bounded multiplicative scaling in the magnitude spectrum for embedding and lightweight linear decoding based on low-level audio representations, the model achieves state-of-the-art robustness against potent attacks, while maintaining high imperceptibility.
Critical strengths
-
This work challenges the conventional belief that strong robustness requires complex latent-space operations or training-time attack simulations, demonstrating that a simple linear scaling predictor and a linear detection head can outperform existing sophisticated baselines.
-
The authors evaluate the watermarking model under extensive realistic attack scenarios, including 18 traditional editing attacks, 30 neural codecs, 7 zero-shot voice cloning, and 4 AI resampling attacks.
Critical weaknesses
-
SimpleMark exhibits substantial performance degradation on certain specific codecs, such as Language-Codec (approximately 50%), WavTokenizer (approximately 64%–71%), and SQ-Codec (approximately 52%).
-
Section 4.4 reveals that the watermark learned by SimpleMark may primarily manifest as a fixed frequency pattern shared across time and independent of audio content. The authors do not sufficiently demonstrate that the watermark can withstand adaptive removal attacks specifically designed to exploit its structural characteristics.
-
The methodological novelty is questionable, as the proposed approach appears to be an effective combination of existing designs rather than an entirely new watermarking mechanism.
Questions for authors
-
The description of the training pipeline is oversimplified and lacks crucial details. The paper presents only a single detection loss in Eq. (2) without explaining how the end-to-end gradient flows from the Detector back through the frozen audio representation backbone, the iSTFT/STFT operators, and into the Embedder's Scale Predictor. It remains unclear whether the Embedder and Detector are trained jointly or in alternative stages, and how numerical stability/gradients are maintained through complex magnitude scaling and phase recombination during iSTFT.
-
Section 4.4 reveals that the watermark learned by SimpleMark may primarily manifest as a fixed frequency pattern shared across time and independent of audio content. The authors do not sufficiently demonstrate that the watermark can withstand adaptive removal attacks specifically designed to exploit its structural characteristics. For example, can an attacker remove the watermark using spectral averaging followed by notch filtering or spectral subtraction while preserving audio quality? Furthermore, could averaging multiple watermarked audio clips enable an attacker to extract the common frequency template and forge ownership claims?
-
In Appendix C.1, the authors adopt a 75% bit-matching threshold for existence classification without explanation. What is the justification for this choice? For , a 75% threshold yields a high False Positive Rate () under random guessing.
3 comments
Question 3: Why is 75% chosen as the threshold used for watermark existence prediction?
As shown in the table below, we choose 75% as the bit-matching threshold since it maximizes mean existence accuracy across four conditions: unattacked audio, codec resynthesis, voice cloning, and AI attacks. We think that this choice balances FPR and FNR under diverse attacks. However, one can also adopt a stricter threshold like 87.50% when FPR matters more. At that threshold, our method can achieve 0.16% FPR with only a 2.06pp decrease in mean existence accuracy.
| Threshold (%) | Theoretical FPR (%) | Mean Empirical FPR (%) | No Atk (%) | Codecs (%) | Cloning (%) | AI (%) | mean (%) |
|---|---|---|---|---|---|---|---|
| 50.00 | 59.82 | 59.93 | 70.73 | 68.30 | 69.50 | 67.72 | 69.06 |
| 56.25 | 40.18 | 40.26 | 80.06 | 76.72 | 79.40 | 75.63 | 77.95 |
| 62.50 | 22.72 | 22.52 | 88.74 | 83.69 | 87.71 | 81.56 | 85.42 |
| 68.75 | 10.51 | 10.33 | 94.92 | 87.13 | 94.09 | 85.44 | 90.40 |
| 75.00 | 3.84 | 3.75 | 98.05 | 87.23 | 96.81 | 86.99 | 92.27 |
| 81.25 | 1.06 | 1.01 | 99.50 | 84.89 | 96.25 | 87.28 | 91.98 |
| 87.50 | 0.21 | 0.17 | 99.94 | 81.08 | 92.87 | 86.95 | 90.21 |
| 93.75 | 0.026 | 0.023 | 99.94 | 75.74 | 84.92 | 85.51 | 86.53 |
| 100.00 | 0.0015 | 0.0009 | 99.27 | 67.26 | 71.95 | 79.48 | 79.49 |
Question 4: Why does SimpleMark perform poorly on Language-Codec, WavTokenizer, and SQ-Codec?
To begin with, it is worth noticing that some codecs are challenging across all evaluated watermarking methods: As shown in the following table, even the strongest baseline suffers from substantial performance degradation on SQ-Codec and WavTokenizer. Therefore, performance degradation on these codecs does not imply that SimpleMark fall behind baselines or is not robust to general neural codecs.
As for the reason why SimpleMark performs poorly on certain codecs, we observed in the following table that recostructed audio quality of these codecs is also comparably low. One plausible explanation is that these codecs heavily alter acoustic details that carry watermark information, making the embedded bits harder to recover.
Nevertheless, we would like to emphasize that SimpleMark achieves substantial gains on several other codecs. For example, compared with the strongest baseline, it improves bit accuracy by 9.52pp on SemantiCodec_0.35 kbps (E3), 6.06pp on SoCodec (M), and 4.66 pp on SemantiCodec_0.70 kbps (E2). On average, SimpleMark achieves the highest mean bit accuracy under neural codec resynthesis attacks, indicating its superior robustness.
| Codec | Bit Acc. (%) | Best Baseline (%) | PESQ |
|---|---|---|---|
| DAC (B) | 99.75 | 99.94 | 3.809 |
| Spectral Codecs (V) | 99.00 | 99.23 | 2.331 |
| Vocos_encodec_12 (L2) | 98.39 | 98.72 | 3.001 |
| EnCodec-6b24k (D1) | 96.39 | 98.89 | 2.331 |
| … | … | … | … |
| Language-Codec (J) | 50.00 | 91.26 | 1.066 |
| WavTokenizer (I1) | 64.00 | 74.26 | 1.918 |
| WavTokenizer (I2) | 71.41 | 75.99 | 1.993 |
| SQ-Codec (W) | 52.79 | 51.79 | 1.222 |
Question 5: What is the methodological novelty beyond an effective combination of existing watermarking designs?
SimpleMark is not a combination of existing designs. Instead, it is a framework building upon two novel mechanisms: Bounded magnitude scaling for imperceiptibiliy control and low-level audio representation-based watermark detection for robustness. These two mechanisms enable superior robustness and competive imperceptility without mainstream techniques like simulated attacks, latent watermarking and perceptual losses, which builds up the core novelty of this paper.
Our default system comprises a linear watermark embedder and a QWen3-ECAPA-based watermark detector. To the best of our knowledge, this kind of architecture has not been reported in existing literature. Indeed, we show in Tab. 2 that our framework is compatible with some network architectures and representations used in existing audio watermarking models. However, choices these alternative modules hardly affect robustness or imperceiptibility. Finally, we also regard simplicity of our framework as an innovation: We demonstrate that state-of-the-art robust audio watermarking can be achieved with few model parameters (12.6M), training data (10k utterances) and simple model architecture.
Thank you for the reply and the additional experiments. Being simple and effective is the selling point of the design, and it is supported. The rating is increased to 3.
We sincerely appreciate your constructive feedback and insightful comments. Now we are going to respond to all your questions and concerns.
Question 1: What is the detail of the training pipeline?
Thank you for pointing out these ambiguities. We jointly train the watermark embedder and detector in a single stage using the detection loss in Eq. (2) as the only supervision. In each training step, we assign a randomly-initialized ground-truth watermark message m to each audio sample, train the embedder and detector in an end-to-end manner with the target of decoding m correctly from the watermarked audio. Although the audio representation extractor is frozen, gradients with respect to its input waveform (aka the watermarked audio xm) are retained. From xm, gradients backpropagate through the differentiable iSTFT to the watermarked magnitude spectrogram M, then through the magnitude scaling and tanh operations to update the scale predictor S. This gradient flow requires neither differentiating phase P nor dividing by low-energy amplitude bins, so numerical stability naturally holds. Meanwhile, the tanh function helps prevent gradient explosion by limiting the gain applied to each bin. After all, the training pipeline of SimpleMark is deliberately designed to be simple and requires no additional stabilization techniques.
Question 2: Is SimpleMark robust against watermark removal or forgery attacks based on spectral averaging?
Thanks for this insightful question. Indeed, we find that existing anti-cloning audio watermarking models, including SimpleMark, are generally vulnerable under adaptive watermark removal and/or forgery attacks. To the best of our knowledge, this kind of attack has not been reported in other literature. Specifically, we consider two types of black-box attacks:
Adaptive Watermark Removal Attack: In this kind of attack, the attacker wants to remove the watermark from a watermarked audio sample$A$. Let MA′ and PA′ denote the STFT amplitude spectrogram of A, and μt the mean across time frames. The attacked STFT amplitude is produced by dividing the original spectrogram with per-bin average, or formally, MA=MA′/μt(MA′). Then the attacked waveform is produced by A=iSTFT(MA,PA′).
Adaptive Watermark Forgery Attack: In this kind of attack, the attacker wants to transfer the watermark from a watermarked audio sample$A$ to an unwatermarked sample B. Let MA′ and MB denote the STFT amplitude spectrogram of A and B, respectively. The attacked STFT amplitude of B is then produced by MB′=MB⋅μt(MB)μt(MA′).
We introduce three metrics to evaluate whether the attack is effective:
Bitwise Accuracy (%): Bitwise accuracy after attack. Lower indicates more effective removal attacks, whereas higher indicates more effective forgery attacks.
Attack Successful Rate (ASR, %): For removal attacks, ASR is the proportion of attacked samples predicted as unwatermarked after the attack. For forgery, it is the proportion of samples predicted as watermarked after the attack.
PESQ: Attacked audio quality. An attack should not be considered successful if it significantly damages audio quality.
Below we report attack results on both baselines and our proposed system:
The results show that all three systems designed for anti-cloning audio watermarking, namely Timbre, VoiceMark and SimpleMark, are susceptible to both adaptive removal and adaptive forgerty attacks. This indicates that watermarks produced by these systems may share the same nature of time-invariant, content agnostic frequency pattern. However, this kind of attack significantly degrades watermarked audio quality. It is therefore questionable whether this kind of attack should be considered practical in reality.
(Continued in the next comment...)