ECViT: Efficient Convolutional Vision Transformer with Local-Attention and Multi-scale Stages

Vision Transformers (ViTs) have revolutionized computer vision by leveraging self-attention to model long-range dependencies. However, ViTs face challenges such as high computational costs due to the quadratic scaling of self-attention and the requirement of a large amount of training data. To address these limitations, we propose the Efficient Convolutional Vision Transformer (ECViT), a hybrid architecture that effectively combines the strengths of CNNs and Transformers. ECViT introduces inductive biases such as locality and translation invariance, inherent to Convolutional Neural Networks (CNNs) into the Transformer framework by extracting patches from low-level features and enhancing the encoder with convolutional operations. Additionally, it incorporates local-attention and a pyramid structure to enable efficient multi-scale feature extraction and representation. Experimental results demonstrate that ECViT achieves an optimal balance between performance and efficiency, outperforming state-of-the-art models on various image classification tasks while maintaining low computational and storage requirements. ECViT offers an ideal solution for applications that prioritize high efficiency without compromising performance.

Fighting Fires from Space: Leveraging Vision Transformers for Enhanced Wildfire Detection and Characterization

Wildfires are increasing in intensity, frequency, and duration across large parts of the world as a result of anthropogenic climate change. Modern hazard detection and response systems that deal with wildfires are under-equipped for sustained wildfire seasons. Recent work has proved automated wildfire detection using Convolutional Neural Networks (CNNs) trained on satellite imagery are capable of high-accuracy results. However, CNNs are computationally expensive to train and only incorporate local image context. Recently, Vision Transformers (ViTs) have gained popularity for their efficient training and their ability to include both local and global contextual information. In this work, we show that ViT can outperform well-trained and specialized CNNs to detect wildfires on a previously published dataset of LandSat-8 imagery. One of our ViTs outperforms the baseline CNN comparison by 0.92%. However, we find our own implementation of CNN-based UNet to perform best in every category, showing their sustained utility in image tasks. Overall, ViTs are comparably capable in detecting wildfires as CNNs, though well-tuned CNNs are still the best technique for detecting wildfire with our UNet providing an IoU of 93.58%, better than the baseline UNet by some 4.58%.

6G WavesFM: A Foundation Model for Sensing, Communication, and Localization

This paper introduces WavesFM, a novel Wireless Foundation Model (WFM) framework, capable of supporting a wide array of communication, sensing, and localization tasks. Our proposed architecture combines a shared Vision Transformer (ViT) backbone with task-specific multi-layer perceptron (MLP) heads and incorporates Low-Rank Adaptation (LoRA) for parameter-efficient fine-tuning. This design promotes full parameter sharing across tasks, significantly reducing the computational and memory footprint without sacrificing performance. The model processes both image-like wireless modalities, such as spectrograms and channel state information (CSI), and in-phase and quadrature (IQ) signals arranged as orthogonal frequency-division multiplexing (OFDM) resource grids. We demonstrate the strong generalization capabilities of WavesFM through extensive experiments on four downstream tasks: Fifth Generation New Radio (5G NR) positioning; multiple-input multiple-output OFDM (MIMO-OFDM) channel estimation; human activity sensing; and radio-frequency (RF) signal classification. Compared to supervised baselines trained individually, our approach achieves superior performance while sharing 80% of its parameters across tasks. Furthermore, we show that pretraining on domain-relevant data not only boosts performance but also accelerates convergence, reducing training time by up to 5x. These results demonstrate that our unified WFM can support diverse tasks and deliver significant gains in both performance and efficiency, highlighting the transformative potential of foundation models to drive AI-native paradigms in future sixth-generation (6G) networks.

Fairness and Robustness in Machine Unlearning

Machine unlearning poses the challenge of ``how to eliminate the influence of specific data from a pretrained model'' in regard to privacy concerns. While prior research on approximated unlearning has demonstrated accuracy and efficiency in time complexity, we claim that it falls short of achieving exact unlearning, and we are the first to focus on fairness and robustness in machine unlearning algorithms. Our study presents fairness Conjectures for a well-trained model, based on the variance-bias trade-off characteristic, and considers their relevance to robustness. Our Conjectures are supported by experiments conducted on the two most widely used model architectures, ResNet and ViT, demonstrating the correlation between fairness and robustness: \textit{the higher fairness-gap is, the more the model is sensitive and vulnerable}. In addition, our experiments demonstrate the vulnerability of current state-of-the-art approximated unlearning algorithms to adversarial attacks, where their unlearned models suffer a significant drop in accuracy compared to the exact-unlearned models. We claim that our fairness-gap measurement and robustness metric should be used to evaluate the unlearning algorithm. Furthermore, we demonstrate that unlearning in the intermediate and last layers is sufficient and cost-effective for time and memory complexity.

RadMamba: Efficient Human Activity Recognition through Radar-based Micro-Doppler-Oriented Mamba State-Space Model

Radar-based HAR has emerged as a promising alternative to conventional monitoring approaches, such as wearable devices and camera-based systems, due to its unique privacy preservation and robustness advantages. However, existing solutions based on convolutional and recurrent neural networks, although effective, are computationally demanding during deployment. This limits their applicability in scenarios with constrained resources or those requiring multiple sensors. Advanced architectures, such as ViT and SSM architectures, offer improved modeling capabilities and have made efforts toward lightweight designs. However, their computational complexity remains relatively high. To leverage the strengths of transformer architectures while simultaneously enhancing accuracy and reducing computational complexity, this paper introduces RadMamba, a parameter-efficient, radar micro-Doppler-oriented Mamba SSM specifically tailored for radar-based HAR. Across three diverse datasets, RadMamba matches the top-performing previous model's 99.8% classification accuracy on Dataset DIAT with only 1/400 of its parameters and equals the leading models' 92.0% accuracy on Dataset CI4R with merely 1/10 of their parameters. In scenarios with continuous sequences of actions evaluated on Dataset UoG2020, RadMamba surpasses other models with significantly higher parameter counts by at least 3%, achieving this with only 6.7k parameters. Our code is available at: https://github.com/lab-emi/AIRHAR.

The Sword of Damocles in ViTs: Computational Redundancy Amplifies Adversarial Transferability

Vision Transformers (ViTs) have demonstrated impressive performance across a range of applications, including many safety-critical tasks. However, their unique architectural properties raise new challenges and opportunities in adversarial robustness. In particular, we observe that adversarial examples crafted on ViTs exhibit higher transferability compared to those crafted on CNNs, suggesting that ViTs contain structural characteristics favorable for transferable attacks. In this work, we investigate the role of computational redundancy in ViTs and its impact on adversarial transferability. Unlike prior studies that aim to reduce computation for efficiency, we propose to exploit this redundancy to improve the quality and transferability of adversarial examples. Through a detailed analysis, we identify two forms of redundancy, including the data-level and model-level, that can be harnessed to amplify attack effectiveness. Building on this insight, we design a suite of techniques, including attention sparsity manipulation, attention head permutation, clean token regularization, ghost MoE diversification, and test-time adversarial training. Extensive experiments on the ImageNet-1k dataset validate the effectiveness of our approach, showing that our methods significantly outperform existing baselines in both transferability and generality across diverse model architectures.

GFT: Gradient Focal Transformer

Fine-Grained Image Classification (FGIC) remains a complex task in computer vision, as it requires models to distinguish between categories with subtle localized visual differences. Well-studied CNN-based models, while strong in local feature extraction, often fail to capture the global context required for fine-grained recognition, while more recent ViT-backboned models address FGIC with attention-driven mechanisms but lack the ability to adaptively focus on truly discriminative regions. TransFG and other ViT-based extensions introduced part-aware token selection to enhance attention localization, yet they still struggle with computational efficiency, attention region selection flexibility, and detail-focus narrative in complex environments. This paper introduces GFT (Gradient Focal Transformer), a new ViT-derived framework created for FGIC tasks. GFT integrates the Gradient Attention Learning Alignment (GALA) mechanism to dynamically prioritize class-discriminative features by analyzing attention gradient flow. Coupled with a Progressive Patch Selection (PPS) strategy, the model progressively filters out less informative regions, reducing computational overhead while enhancing sensitivity to fine details. GFT achieves SOTA accuracy on FGVC Aircraft, Food-101, and COCO datasets with 93M parameters, outperforming ViT-based advanced FGIC models in efficiency. By bridging global context and localized detail extraction, GFT sets a new benchmark in fine-grained recognition, offering interpretable solutions for real-world deployment scenarios.

Change State Space Models for Remote Sensing Change Detection

Despite their frequent use for change detection, both ConvNets and Vision transformers (ViT) exhibit well-known limitations, namely the former struggle to model long-range dependencies while the latter are computationally inefficient, rendering them challenging to train on large-scale datasets. Vision Mamba, an architecture based on State Space Models has emerged as an alternative addressing the aforementioned deficiencies and has been already applied to remote sensing change detection, though mostly as a feature extracting backbone. In this article the Change State Space Model is introduced, that has been specifically designed for change detection by focusing on the relevant changes between bi-temporal images, effectively filtering out irrelevant information. By concentrating solely on the changed features, the number of network parameters is reduced, enhancing significantly computational efficiency while maintaining high detection performance and robustness against input degradation. The proposed model has been evaluated via three benchmark datasets, where it outperformed ConvNets, ViTs, and Mamba-based counterparts at a fraction of their computational complexity. The implementation will be made available at https://github.com/Elman295/CSSM upon acceptance.

VEXP: A Low-Cost RISC-V ISA Extension for Accelerated Softmax Computation in Transformers

While Transformers are dominated by Floating-Point (FP) Matrix-Multiplications, their aggressive acceleration through dedicated hardware or many-core programmable systems has shifted the performance bottleneck to non-linear functions like Softmax. Accelerating Softmax is challenging due to its non-pointwise, non-linear nature, with exponentiation as the most demanding step. To address this, we design a custom arithmetic block for Bfloat16 exponentiation leveraging a novel approximation algorithm based on Schraudolph's method, and we integrate it into the Floating-Point Unit (FPU) of the RISC-V cores of a compute cluster, through custom Instruction Set Architecture (ISA) extensions, with a negligible area overhead of 1\%. By optimizing the software kernels to leverage the extension, we execute Softmax with 162.7$\times$ less latency and 74.3$\times$ less energy compared to the baseline cluster, achieving an 8.2$\times$ performance improvement and 4.1$\times$ higher energy efficiency for the FlashAttention-2 kernel in GPT-2 configuration. Moreover, the proposed approach enables a multi-cluster system to efficiently execute end-to-end inference of pre-trained Transformer models, such as GPT-2, GPT-3 and ViT, achieving up to 5.8$\times$ and 3.6$\times$ reduction in latency and energy consumption, respectively, without requiring re-training and with negligible accuracy loss.

DMPT: Decoupled Modality-aware Prompt Tuning for Multi-modal Object Re-identification

Current multi-modal object re-identification approaches based on large-scale pre-trained backbones (i.e., ViT) have displayed remarkable progress and achieved excellent performance. However, these methods usually adopt the standard full fine-tuning paradigm, which requires the optimization of considerable backbone parameters, causing extensive computational and storage requirements. In this work, we propose an efficient prompt-tuning framework tailored for multi-modal object re-identification, dubbed DMPT, which freezes the main backbone and only optimizes several newly added decoupled modality-aware parameters. Specifically, we explicitly decouple the visual prompts into modality-specific prompts which leverage prior modality knowledge from a powerful text encoder and modality-independent semantic prompts which extract semantic information from multi-modal inputs, such as visible, near-infrared, and thermal-infrared. Built upon the extracted features, we further design a Prompt Inverse Bind (PromptIBind) strategy that employs bind prompts as a medium to connect the semantic prompt tokens of different modalities and facilitates the exchange of complementary multi-modal information, boosting final re-identification results. Experimental results on multiple common benchmarks demonstrate that our DMPT can achieve competitive results to existing state-of-the-art methods while requiring only 6.5% fine-tuning of the backbone parameters.