arXiv subjects

Jilei Sun

Publications and source records attributed to Jilei Sun.

4 recordsLinked to original sources

arXiv2026

Tilewise Domain-Separated Selective Encryption for Remote Sensing Imagery under Chosen-Plaintext Attacks

Selective image encryption is common in remote sensing systems because it protects sensitive regions of interest (ROI) while limiting computational cost. However, many selective designs enable cross-tile structural leakage under chosen-plaintext attacks when secret-dependent transformations are reused across spatial positions. This paper proposes Tilewise Domain-Separated Selective Encryption (TDS-SE), where per-tile (and optionally per-frame) keys are derived from a master secret via HKDF with explicit domain separation, and ROI masks are treated strictly as external side information. Structural leakage is evaluated using two reconstruction-based distinguishers -- a linear model and a lightweight convolutional neural network -- under multiple attack settings. Experiments on RESISC45 and SEN12MS cover ablation tests, cross-position transferability, cross-sample generalization, and ROI-knowledge asymmetry. Results show that per-tile domain separation reduces position-conditioned transfer for the linear probe, and that adding frame freshness improves robustness to imperfect ROI assumptions for the CNN probe. Cross-sample generalization exhibits mixed behavior across settings, consistent with an empirical evaluation perspective, while selective-encryption functionality is preserved under the same tiling and ROI policy. Beyond the method itself, the paper provides a structured protocol for evaluating selective encryption under realistic attacker capabilities.

cs.CR↗

arXiv2025

Can Video Large Multimodal Models Think Like Doubters-or Double-Down: A Study on Defeasible Video Entailment

Video Large Multimodal Models (VLMMs) have made impressive strides in understanding video content, but they often struggle with abstract and adaptive reasoning-the ability to revise their interpretations when new information emerges. In reality, conclusions are rarely set in stone; additional context can strengthen or weaken an initial inference. To address this, we introduce Defeasible Video Entailment (DVidE), a new task that challenges models to think like doubters, constantly updating their reasoning based on evolving evidence. In DVidE, given a video premise and a textual hypothesis, models must determine whether a new update strengthens or weakens the hypothesis (classification version) or generate a coherent update that modifies the entailment relationship (generation version). For solving the classification task, we propose the Chain of Counterfactual Thought framework, utilizing counterfactual reasoning, ASR-enhanced video content, and rationale refinement to reduce inference bias. For the generation task, we develop a framework that combines ASR output with a Large Language Model (LLM) to produce coherent, contextually relevant updates aligned with the intended strengthener or weakener goals. Additionally, we introduce a novel benchmark dataset, with strengthener/weakener annotations and an LLM-based evaluation metric specifically designed for assessing generative performance. Experimental results demonstrate significant improvements, highlighting our proposed method in enhancing dynamic reasoning capabilities of VLMMs.

cs.CV↗

arXiv2025

A 3D-Cascading Crossing Coupling Framework for Hyperchaotic Map Construction and Its Application to Color Image Encryption

This paper focuses on hyperchaotic-map construction and proposes a 3D-Cascading Crossing Coupling framework (3D-CCC), which cascades, crosses, and couples three one-dimensional chaotic maps to form a three-dimensional hyperchaotic system. The framework avoids modulo-1 operations and introduces bounded-state and denominator safeguards for stable digital implementation. A general 3D-CCC formulation is established, and its derivative/Jacobian structure is analyzed to characterize multidirectional expansion. By instantiating ICMIC, Logistic, and Sine maps, a concrete system (3D-ILS) is derived. Phase portraits, bifurcation behavior, sensitivity tests, and Lyapunov-exponent analysis indicate pronounced ergodicity and hyperchaotic dynamics. As an application of the constructed map, a one-round RGB image-encryption scheme is developed using cross-channel bit mixing with joint permutation-diffusion. Under the reported settings, the cipher reaches near-ideal entropy (average 7.9993), NPCR of 96.61\%, UACI of 33.46\%, and an effective key space of about $2^{309}$. These results support the effectiveness of 3D-CCC as a practical framework for hyperchaotic-system design, with image encryption as one representative application.

cs.MM↗

arXiv2021

Extension of the operation mode of back-streaming white neutron beam at the China spallation neutron source

The back-streaming white neutron source (Back-n) is a comprehensive neutron experimental platform for nuclear data measurement, nuclear astrophysics, neutron irradiation, detector calibration, etc. In order to meet a variety of experimental requirements, an attempt of various combinations of collimators has been made for the current collimation system. The basic parameters such as beam flux and beam spot characteristics under different beam operation modes have been studied. According to the change of the CSNS proton beam operation mode, the influence of different proton beam spots on the Back-n beam is also studied. The study finds that some beam modes with new collimator combinations can meet the experimental requirements and greatly shorten the experimental time. In addition, we have found a parameterized formula for the uniform beam spot of the high-power proton beam based on the beam monitor data, which provides a more accurate proton-beam source term for future white neutron beam physics and application researches.

physics.ins-det↗