Tomoaki Ohtsuki

Lingfeng Yao, Xincong Zhong, Chenpei Huang et al.

With the rise of AI-generated audio, watermarking has become widely used for detecting misuse and protecting intellectual property. However, adversaries may try to remove these watermarks, making it critical to evaluate how well watermarking schemes withstand removal attacks. Existing attacks are often impractical: they either noticeably degrade perceptual quality or require access to the watermarking scheme. We propose DiffErase, a black-box watermark removal attack that assumes no knowledge of the target watermarking scheme while maintaining perceptual quality. DiffErase perturbs watermarked audio to an intermediate diffusion noise level and regenerates it using a pretrained denoising model, effectively suppressing watermark signals. Theoretical analysis and extensive experiments demonstrate that inaudible audio watermarks are highly vulnerable: across multiple audio domains, DiffErase consistently removes watermarks while preserving perceptual quality. These findings highlight the need for future audio watermarking designs to consider diffusion-based threats. Code and demos are available at https://differase.github.io/DiffErase/.

Yusaku Negoya, Feifei Cui, Zilong Zhang et al.

Omics data is widely employed in medical research to identify disease mechanisms and contains highly sensitive personal information. Federated Learning (FL) with Differential Privacy (DP) can ensure the protection of omics data privacy against malicious user attacks. However, FL with the DP method faces an inherent trade-off: stronger privacy protection degrades predictive accuracy due to injected noise. On the other hand, Homomorphic Encryption (HE) allows computations on encrypted data and enables aggregation of encrypted gradients without DP-induced noise can increase the predictive accuracy. However, it may increase the computation cost. To improve the predictive accuracy while considering the computational ability of heterogeneous clients, we propose a Privacy-Preserving Machine Learning (PPML)-Hybrid method by introducing HE. In the proposed PPML-Hybrid method, clients distributed select either HE or DP based on their computational resources, so that HE clients contribute noise-free updates while DP clients reduce computational overhead. Meanwhile, clients with high computational resources clients can flexibly adopt HE or DP according to their privacy needs. Performance evaluation on omics datasets show that our proposed method achieves comparable predictive accuracy while significantly reducing computation time relative to HE-only. Additionally, it outperforms DP-only methods under equivalent or stricter privacy budgets.

Zhangcheng Hou, Tomoaki Ohtsuki

Radar-camera depth estimation must turn an ultra-sparse, all-weather, metric radar signal into a dense per-pixel depth map. Existing methods -- concatenation, confidence-aware gating, sparse supervision, graph-based extraction -- combine radar and image features outside the backbone's sequence operator, and even cross-modal Mamba variants leave the selection mechanism itself unimodal. We argue that the selection mechanism is the right place for radar to enter. We introduce Radar-Modulated Selection (RMS), a minimal and principled way to inject radar into Mamba's selective scan: radar modulates the scan from within, adding zero-initialised perturbations to the step size $Δ$ and readout $\mathbf{C}$ while leaving the input projection $\mathbf{B}$ and state dynamics $\mathbf{A}$ image-only. The construction is exactly equivalent to a pretrained image-only Mamba at initialisation, ensuring radar only influences the model where it improves accuracy. Two further properties follow that out-of-scan fusion cannot offer: linear-cost cross-modal coupling at every recurrence step, and a natural fallback to the image-only backbone when radar is absent. We deploy RMS in a Multi-View Scan Pyramid (MVSP) that matches the fusion operator to radar's spatial reach at each scale. SemoDepth achieves state-of-the-art performance on nuScenes, reducing MAE by 34.0%, 29.9%, and 29.9% over the previous best at 0--50, 0--70, and 0--80m, while attaining the lowest single-frame latency (26.8ms). A further ablation shows that out-of-scan feature blending adds no accuracy on top of RMS, providing empirical validation that in-scan selection can replace out-of-scan fusion.

Liang Li, Xingke Yang, Wen Wu et al.

Large Language Model (LLM) at mobile devices and its potential applications never fail to fascinate. However, on-device LLM fine-tuning poses great challenges due to extremely high memory requirements and slow training speeds. Even with parameter-efficient fine-tuning (PEFT) methods that update only a small subset of parameters, resource-constrained mobile devices cannot afford them. In this paper, we propose MobiLLM to enable memory-efficient transformer LLM fine-tuning on a mobile device via server-assisted side-tuning. Particularly, MobiLLM allows the resource-constrained mobile device to retain merely a frozen backbone model, while offloading the memory and computation-intensive backpropagation of a trainable side-network to a high-performance server. Unlike existing fine-tuning methods that keep trainable parameters inside the frozen backbone, MobiLLM separates a set of parallel adapters from the backbone to create a backpropagation bypass, involving only one-way activation transfers from the mobile device to the server with low-width quantization during forward propagation. In this way, the data never leaves the mobile device while the device can remove backpropagation through the local backbone model and its forward propagation can be paralyzed with the server-side execution. Thus, MobiLLM preserves data privacy while significantly reducing the memory and computational burdens for LLM fine-tuning. Through extensive experiments, we demonstrate that MobiLLM can enable a resource-constrained mobile device, even a CPU-only one, to fine-tune LLMs and significantly reduce convergence time and memory usage.

Xingke Yang, Liang Li, Zhiyi Wan et al.

There is a huge gap between numerous intriguing applications fostered by on-device large language model (LLM) fine-tuning (FT) from fresh mobile data and the limited resources of a mobile device. While existing server-assisted methods (e.g., split learning or side-tuning) may enable LLM FT on the local mobile device, they suffer from heavy communication burdens of activation transmissions, and may disclose data and labels to the server. To address those issues, we develop PAE MobiLLM, a a privacy-aware and efficient LLM FT method which can be deployed on the mobile device via server-assisted additive side-tuning. To further accelerate FT convergence and improve computing efficiency, PAE MobiLLM integrates activation caching on the server side, which allows the server to reuse historical activations and saves the mobile device from repeatedly computing forward passes for the recurring data samples. Besides, to reduce communication cost, PAE MobiLLM develops an activation shortcut that transmits only the token involved in the loss calculation instead of full activation matrices to guide the side network tuning. Last but not least, PAE MobiLLM introduces the additive adapter side-network design which makes the server train the adapter modules based on device-defined prediction differences rather than raw ground-truth labels. In this way, the server can only assist device-defined side-network computing, and learn nothing about data and labels. Extensive experimental results demonstrate PAE MobiLLM's superiority.

Guijun Liu, Yuwen Cao, Tomoaki Ohtsuki et al.

Deep autoencoder (DAE) frameworks have demonstrated their effectiveness in reducing channel state information (CSI) feedback overhead in massive multiple-input multiple-output (mMIMO) orthogonal frequency division multiplexing (OFDM) systems. However, existing CSI feedback models struggle to adapt to dynamic environments caused by user mobility, requiring retraining when encountering new CSI distributions. Moreover, returning to previously encountered environments often leads to performance degradation due to catastrophic forgetting. Continual learning involves enabling models to incorporate new information while maintaining performance on previously learned tasks. To address these challenges, we propose a generative adversarial network (GAN)-based learning approach for CSI feedback. By using a GAN generator as a memory unit, our method preserves knowledge from past environments and ensures consistently high performance across diverse scenarios without forgetting. Simulation results show that the proposed approach enhances the generalization capability of the DAE framework while maintaining low memory overhead. Furthermore, it can be seamlessly integrated with other advanced CSI feedback models, highlighting its robustness and adaptability.

Yuwen Cao, Guijun Liu, Tomoaki Ohtsuki et al.

The deep autoencoder (DAE) framework has turned out to be efficient in reducing the channel state information (CSI) feedback overhead in massive multiple-input multipleoutput (mMIMO) systems. However, these DAE approaches presented in prior works rely heavily on large-scale data collected through the base station (BS) for model training, thus rendering excessive bandwidth usage and data privacy issues, particularly for mMIMO systems. When considering users' mobility and encountering new channel environments, the existing CSI feedback models may often need to be retrained. Returning back to previous environments, however, will make these models perform poorly and face the risk of catastrophic forgetting. To solve the above challenging problems, we propose a novel gossiping generative adversarial network (Gossip-GAN)-aided CSI feedback training framework. Notably, Gossip-GAN enables the CSI feedback training with low-overhead while preserving users' privacy. Specially, each user collects a small amount of data to train a GAN model. Meanwhile, a fully distributed gossip-learning strategy is exploited to avoid model overfitting, and to accelerate the model training as well. Simulation results demonstrate that Gossip-GAN can i) achieve a similar CSI feedback accuracy as centralized training with real-world datasets, ii) address catastrophic forgetting challenges in mobile scenarios, and iii) greatly reduce the uplink bandwidth usage. Besides, our results show that the proposed approach possesses an inherent robustness.

Yuwen Cao, Tomoaki Ohtsuki, Setareh Maghsudi et al.

In this paper, we develop a deep learning (DL)-guided hybrid beam and power allocation approach for multiuser millimeter-wave (mmWave) networks, which facilitates swift beamforming at the base station (BS). The following persisting challenges motivated our research: (i) User and vehicular mobility, as well as redundant beam-reselections in mmWave networks, degrade the efficiency; (ii) Due to the large beamforming dimension at the BS, the beamforming weights predicted by the cutting-edge DL-based methods often do not suit the channel distributions; (iii) Co-located user devices may cause a severe beam conflict, thus deteriorating system performance. To address the aforementioned challenges, we exploit the synergy of supervised learning and super-resolution technology to enable low-overhead beam- and power allocation. In the first step, we propose a method for beam-quality prediction. It is based on deep learning and explores the relationship between high- and low-resolution beam images (energy). Afterward, we develop a DL-based allocation approach, which enables high-accuracy beam and power allocation with only a portion of the available time-sequential low-resolution images. Theoretical and numerical results verify the effectiveness of our proposed