Search Results for author:
Found 13 papers, 8 papers with code
Eliminating Information Leakage in Hard Concept Bottleneck Models with Supervised, Hierarchical Concept Learning
This paper alleviates the information leakage issue by introducing label supervision in concept predication and constructing a hierarchical concept set.
No Privacy Left Outside: On the (In-)Security of TEE-Shielded DNN Partition for On-Device ML
These solutions, referred to as TEE-Shielded DNN Partition (TSDP), partition a DNN model into two parts, offloading the privacy-insensitive part to the GPU while shielding the privacy-sensitive part within the TEE.
Unveiling Single-Bit-Flip Attacks on DNN Executables
Nevertheless, we find that DNN executables contain extensive, severe (e. g., single-bit flip), and transferrable attack surfaces that are not present in high-level DNN models and can be exploited to deplete full model intelligence and control output labels.
Precise and Generalized Robustness Certification for Neural Networks
We identify two key properties, independence and continuity, that convert the latent space into a precise and analysis-friendly input space representation for certification.
Explain Any Concept: Segment Anything Meets Concept-Based Explanation
For computer vision tasks, mainstream pixel-based XAI methods explain DNN decisions by identifying important pixels, and emerging concept-based XAI explore forming explanations with concepts (e. g., a head in an image).
Decompiling x86 Deep Neural Network Executables
BTD takes DNN executables and outputs full model specifications, including types of DNN operators, network topology, dimensions, and parameters that are (nearly) identical to those of the input models.
ADI: Adversarial Dominating Inputs in Vertical Federated Learning Systems
Vertical federated learning (VFL) system has recently become prominent as a concept to process data distributed across many individual sources without the need to centralize it.
Automated Side Channel Analysis of Media Software with Manifold Learning
Recent advances in representation learning and perceptual learning inspired us to consider the reconstruction of media inputs from side channel traces as a cross-modality manifold learning task that can be addressed in a unified manner with an autoencoder framework trained to learn the mapping between media inputs and side channel observations.
MDPFuzz: Testing Models Solving Markov Decision Processes
During fuzzing, MDPFuzz decides which mutated state to retain by measuring if it can reduce cumulative rewards or form a new state sequence.
Provably Valid and Diverse Mutations of Real-World Media Data for DNN Testing
In contrast, we discuss the feasibility of mutating real-world media data with provably high DIV and VAL based on manifold.
Revisiting Neuron Coverage for DNN Testing: A Layer-Wise and Distribution-Aware Criterion
We demonstrate that NLC is significantly correlated with the diversity of a test suite across a number of tasks (classification and generation) and data formats (image and text).
Perception Matters: Detecting Perception Failures of VQA Models Using Metamorphic Testing
MetaVQA checks whether the answer to (i, q) satisfies metamorphic relationships (MRs), denoting perception consistency, with the composed answers of transformed questions and images.
Private Image Reconstruction from System Side Channels Using Generative Models
Given the ever-growing adoption of machine learning as a service (MLaaS), image analysis software on cloud platforms has been exploited by reconstructing private user images from system side channels.
