arXiv subjects

M. Wu

Publications and source records attributed to M. Wu.

At least 19 recordsLinked to original sources

arXiv2026

Security Evaluation of Laser-Phase-Noise Quantum Random Number Generators with Intrinsic Correlations

Quantum random number generators are essential for achieving information-theoretical security in modern cryptographic systems. Among various implementations, laser phase noise schemes are widely favored for their simple architecture and high integration potential. However, the intrinsic correlations in the raw data are often neglected, which violates the independent and identically distributed assumption and potentially compromises system security. In this work, we establish an analytical model of correlation and formulate an analytical expression for the conditional min-entropy in the presence of intrinsic correlations to accurately quantify the genuinely extractable randomness. The validity of our theoretical model is confirmed by numerical simulations and experimental results, exhibiting excellent agreement. Under typical setups it is shown that neglecting intrinsic correlations leads to an overestimation of extractable randomness by approximately 46%. This work provides a valuable theoretical framework for designing compact, high-performance quantum random number generators with rigorous security analysis.

quant-ph↗

arXiv2026

Enhancing Train-Free Infinite-Frame Generation for Consistent Long Videos

Without incurring significant computational overhead, train-free long video generation aims to enable foundation video generation models to produce longer videos. Frame-level autoregressive frameworks, e.g., FIFO-diffusion, offer the advantage of generating infinitely long videos with constant memory consumption. However, the mismatch between training and inference, coupled with the challenge of maintaining long-term consistency, limits the effective utilization of foundation models. To mitigate these concerns, we propose \textbf{MIGA}, a novel infinite-frame long video generation method. Firstly, we propose an effective two-stage alignment mechanism that mitigates the training-inference gap by reducing the excessive noise span fed to the model. We then introduce an innovative dual consistency enhancement mechanism, where the self-reflection approach corrects early high-noise frames and the long-range frame guidance approach leverages later low-noise frames with broad coverage to steer generation, jointly improving temporal consistency. Extensive experiments on VBench and NarrLV demonstrate the state-of-the-art performance of MIGA. Our project page is available at https://xiaokunfeng.github.io/miga_homepage/.

cs.CV↗

arXiv2025

ATCTrack: Aligning Target-Context Cues with Dynamic Target States for Robust Vision-Language Tracking

Vision-language tracking aims to locate the target object in the video sequence using a template patch and a language description provided in the initial frame. To achieve robust tracking, especially in complex long-term scenarios that reflect real-world conditions as recently highlighted by MGIT, it is essential not only to characterize the target features but also to utilize the context features related to the target. However, the visual and textual target-context cues derived from the initial prompts generally align only with the initial target state. Due to their dynamic nature, target states are constantly changing, particularly in complex long-term sequences. It is intractable for these cues to continuously guide Vision-Language Trackers (VLTs). Furthermore, for the text prompts with diverse expressions, our experiments reveal that existing VLTs struggle to discern which words pertain to the target or the context, complicating the utilization of textual cues. In this work, we present a novel tracker named ATCTrack, which can obtain multimodal cues Aligned with the dynamic target states through comprehensive Target-Context feature modeling, thereby achieving robust tracking. Specifically, (1) for the visual modality, we propose an effective temporal visual target-context modeling approach that provides the tracker with timely visual cues. (2) For the textual modality, we achieve precise target words identification solely based on textual content, and design an innovative context words calibration method to adaptively utilize auxiliary context words. (3) We conduct extensive experiments on mainstream benchmarks and ATCTrack achieves a new SOTA performance. The code and models will be released at: https://github.com/XiaokunFeng/ATCTrack.

cs.CV↗

arXiv2025

NarrLV: Towards a Comprehensive Narrative-Centric Evaluation for Long Video Generation

With the rapid development of foundation video generation technologies, long video generation models have exhibited promising research potential thanks to expanded content creation space. Recent studies reveal that the goal of long video generation tasks is not only to extend video duration but also to accurately express richer narrative content within longer videos. However, due to the lack of evaluation benchmarks specifically designed for long video generation models, the current assessment of these models primarily relies on benchmarks with simple narrative prompts (e.g., VBench). To the best of our knowledge, our proposed NarrLV is the first benchmark to comprehensively evaluate the Narrative expression capabilities of Long Video generation models. Inspired by film narrative theory, (i) we first introduce the basic narrative unit maintaining continuous visual presentation in videos as Temporal Narrative Atom (TNA), and use its count to quantitatively measure narrative richness. Guided by three key film narrative elements influencing TNA changes, we construct an automatic prompt generation pipeline capable of producing evaluation prompts with a flexibly expandable number of TNAs. (ii) Then, based on the three progressive levels of narrative content expression, we design an effective evaluation metric using the MLLM-based question generation and answering framework. (iii) Finally, we conduct extensive evaluations on existing long video generation models and the foundation generation models. Experimental results demonstrate that our metric aligns closely with human judgments. The derived evaluation outcomes reveal the detailed capability boundaries of current video generation models in narrative content expression.

cs.CV↗

arXiv2025

CSTrack: Enhancing RGB-X Tracking via Compact Spatiotemporal Features

Effectively modeling and utilizing spatiotemporal features from RGB and other modalities (\eg, depth, thermal, and event data, denoted as X) is the core of RGB-X tracker design. Existing methods often employ two parallel branches to separately process the RGB and X input streams, requiring the model to simultaneously handle two dispersed feature spaces, which complicates both the model structure and computation process. More critically, intra-modality spatial modeling within each dispersed space incurs substantial computational overhead, limiting resources for inter-modality spatial modeling and temporal modeling. To address this, we propose a novel tracker, CSTrack, which focuses on modeling Compact Spatiotemporal features to achieve simple yet effective tracking. Specifically, we first introduce an innovative Spatial Compact Module that integrates the RGB-X dual input streams into a compact spatial feature, enabling thorough intra- and inter-modality spatial modeling. Additionally, we design an efficient Temporal Compact Module that compactly represents temporal features by constructing the refined target distribution heatmap. Extensive experiments validate the effectiveness of our compact spatiotemporal modeling method, with CSTrack achieving new SOTA results on mainstream RGB-X benchmarks. The code and models will be released at: https://github.com/XiaokunFeng/CSTrack.

cs.CV↗

arXiv2025

New insight into the Rapid Burster by Insight-HXMT

We report the timing and spectral analyses upon of the type II X-ray bursts from the Rapid Burster (MXB 1730--335) observed by Insight-HXMT and Swift/XRT. By stacking the long-duration bursts, we find for the first time that the hard X-rays are lagging than the soft X-rays by 3 seconds. However, such a lag is not visible for the short-duration bursts, probably because of the poor statistics. For all bursts the energy spectrum is found to be non-thermal, thanks to the broad band coverage of Insight-HXMT. These findings put new insights into the type-II bursts and require a temporally showing-up corona for possible interpretation.

astro-ph.HE↗

arXiv2024

Enhancing Vision-Language Tracking by Effectively Converting Textual Cues into Visual Cues

Vision-Language Tracking (VLT) aims to localize a target in video sequences using a visual template and language description. While textual cues enhance tracking potential, current datasets typically contain much more image data than text, limiting the ability of VLT methods to align the two modalities effectively. To address this imbalance, we propose a novel plug-and-play method named CTVLT that leverages the strong text-image alignment capabilities of foundation grounding models. CTVLT converts textual cues into interpretable visual heatmaps, which are easier for trackers to process. Specifically, we design a textual cue mapping module that transforms textual cues into target distribution heatmaps, visually representing the location described by the text. Additionally, the heatmap guidance module fuses these heatmaps with the search image to guide tracking more effectively. Extensive experiments on mainstream benchmarks demonstrate the effectiveness of our approach, achieving state-of-the-art performance and validating the utility of our method for enhanced VLT.

cs.CV↗

arXiv2022

Peculiar disk behaviors of the black hole candidate MAXI J1348-630 in the hard state observed by Insight-HXMT and Swift

We present a spectral study of the black hole candidate MAXI J1348-630 during its 2019 outburst, based on monitoring observations with Insight-HXMT and Swift. Throughout the outburst, the spectra are well fitted with power-law plus disk-blackbody components. In the soft-intermediate and soft states, we observed the canonical relation L ~ T_in^4 between disk luminosity L and peak colour temperature T_in, with a constant inner radius R_in (traditionally identified with the innermost stable circular orbit). At other stages of the outburst cycle, the behaviour is more unusual, inconsistent with the canonical outburst evolution of black hole transients. In particular, during the hard rise, the apparent inner radius is smaller than in the soft state (and increasing), and the peak colour temperature is higher (and decreasing). This anomalous behaviour is found even when we model the spectra with self-consistent Comptonization models, which take into account the up-scattering of photons from the disk component into the power-law component. To explain both those anomalous trends at the same time, we suggest that the hardening factor for the inner disk emission was larger than the canonical value of ~1.7 at the beginning of the outburst. A more physical trend of radii and temperature evolution requires a hardening factor evolving from ~3.5 at the beginning of the hard state to ~1.7 in the hard intermediate state. This could be evidence that the inner disk was in the process of condensing from the hot, optically thin medium and had not yet reached a sufficiently high optical depth for its emission spectrum to be described by the standard optically-thick disk solution.

astro-ph.HE↗

arXiv2021

Roadmap on Spin-Wave Computing

Magnonics is a field of science that addresses the physical properties of spin waves and utilizes them for data processing. Scalability down to atomic dimensions, operations in the GHz-to-THz frequency range, utilization of nonlinear and nonreciprocal phenomena, and compatibility with CMOS are just a few of many advantages offered by magnons. Although magnonics is still primarily positioned in the academic domain, the scientific and technological challenges of the field are being extensively investigated, and many proof-of-concept prototypes have already been realized in laboratories. This roadmap is a product of the collective work of many authors that covers versatile spin-wave computing approaches, conceptual building blocks, and underlying physical phenomena. In particular, the roadmap discusses the computation operations with Boolean digital data, unconventional approaches like neuromorphic computing, and the progress towards magnon-based quantum computing. The article is organized as a collection of sub-sections grouped into seven large thematic sections. Each sub-section is prepared by one or a group of authors and concludes with a brief description of the current challenges and the outlook of the further development of the research directions.

physics.app-ph↗

arXiv2021

Search for Gamma-Ray Bursts and Gravitational Wave Electromagnetic Counterparts with High Energy X-ray Telescope of \textit{Insight}-HXMT

The High Energy X-ray telescope (HE) on-board the Hard X-ray Modulation Telescope (\textit{Insight}-HXMT) can serve as a wide Field of View (FOV) gamma-ray monitor with high time resolution ($\mu$s) and large effective area (up to thousands cm$^2$). We developed a pipeline to search for Gamma-Ray Bursts (GRBs), using the traditional signal-to-noise ratio (SNR) method for blind search and the coherent search method for targeted search. By taking into account the location and spectrum of the burst and the detector response, the targeted coherent search is more powerful to unveil weak and sub-threshold bursts, especially those in temporal coincidence with Gravitational Wave (GW) events. Based on the original method in literature, we further improved the coherent search to filter out false triggers caused by spikes in light curves, which are commonly seen in gamma-ray instruments (e.g. \textit{Fermi}/GBM, \textit{POLAR}). We show that our improved targeted coherent search method could eliminate almost all false triggers caused by spikes. Based on the first two years of \textit{Insight}-HXMT/HE data, our targeted search recovered 40 GRBs, which were detected by either \textit{Swift}/BAT or \textit{Fermi}/GBM but too weak to be found in our blind search. With this coherent search pipeline, the GRB detection sensitivity of \textit{Insight}-HXMT/HE is increased to about 1.5E-08 erg/cm$^2$ (200 keV$-$3 MeV). We also used this targeted coherent method to search \textit{Insight}-HXMT/HE data for electromagnetic (EM) counterparts of LIGO-Virgo GW events (including O2 and O3a runs). However, we did not find any significant burst associated with GW events.

astro-ph.HE↗

arXiv2021

Accretion Torque Reversals in GRO J1008-57 Revealed by Insight-HXMT

GRO J1008-57, as a Be/X-ray transient pulsar, is considered to have the highest magnetic field in known neutron star X-ray binary systems. Observational data of the X-ray outbursts in GRO J1008-57 from 2017 to 2020 were collected by the Insight-HXMT satellite. In this work, the spin period of the neutron star in GRO J1008-57 was determined to be about 93.28 seconds in August 2017, 93.22 seconds in February 2018, 93.25 seconds in June 2019 and 93.14 seconds in June 2020. GRO J1008-57 evolved in the spin-up process with a mean rate of $-(2.10\pm 0.05)\times$10$^{-4}$ s/d from 2009 -- 2018, and turned into a spin down process with a rate of $(6.7\pm 0.6)\times$10$^{-5}$ s/d from Feb 2018 to June 2019. During the type II outburst of 2020, GRO J1008-57 had the spin-up torque again. During the torque reversals, the pulse profiles and continuum X-ray spectra did not change significantly, and the cyclotron resonant scattering feature around 80 keV was only detected during the outbursts in 2017 and 2020. Based on the observed mean spin-up rate, we estimated the inner accretion disk radius in GRO J1008-57 (about 1 - 2 times of the Alfvén radius) by comparing different accretion torque models of magnetic neutron stars. During the spin-down process, the magnetic torque should dominate over the matter accreting inflow torque, and we constrained the surface dipole magnetic field $B\geq 6\times 10^{12}$ G for the neutron star in GRO J1008-57, which is consistent with the magnetic field strength obtained by cyclotron line centroid energy.

astro-ph.HE↗

arXiv2021

QPOs and Orbital elements of X-ray binary 4U 0115+63 during the 2017 outburst observed by Insight-HXMT

In this paper, we presented a detailed timing analysis of a prominent outburst of 4U 0115+63 detected by \textit{Insight}-HXMT in 2017 August. The spin period of the neutron star was determined to be $3.61398\pm 0.00002$ s at MJD 57978. We measured the period variability and extract the orbital elements of the binary system. The angle of periastron evolved with a rate of $0.048\pm0.003$ $yr^{-1}$. The light curves are folded to sketch the pulse profiles in different energy ranges. A multi-peak structure in 1-10 keV is clearly illustrated. We introduced wavelet analysis into our data analysis procedures to study QPO signals and perform a detailed wavelet analysis in many different energy ranges. Through the wavelet spectra, we report the discovery of a QPO at the frequency $\sim 10$ mHz. In addition, the X-ray light curves showed multiple QPOs in the period of $\sim 16-32 $ s and $\sim 67- 200 $ s. We found that the $\sim100$ s QPO was significant in most of the observations and energies. There exist positive relations between X-ray luminosity and their Q-factors and S-factors, while the QPO periods have no correlation with X-ray luminosity. In wavelet phase maps, we found that the pulse phase of $\sim 67- 200 $ s QPO drifting frequently while the $\sim 16-32 $ s QPO scarcely drifting. The dissipation of oscillations from high energy to low energy was also observed. These features of QPOs in 4U 0115+63 provide new challenge to our understanding of their physical origins.

astro-ph.HE↗

arXiv2020

A variable ionized disk wind in the black-hole candidate EXO 1846-031

After 34 years, the black-hole candidate EXO 1846-031 went into outburst again in 2019. We investigate its spectral properties in the hard intermediate and the soft states with NuSTAR and Insight-HXMT. A reflection component has been detected in the two spectral states but possibly originating from different illumination spectra: in the intermediate state, the illuminating source is attributed to a hard coronal component, which has been commonly observed in other X-ray binaries, whereas in the soft state the reflection is probably produced by the disk self-irradiation. Both cases support EXO 1846-031 as a low inclination system of ~40 degrees. An absorption line is clearly detected at ~7.2 keV in the hard intermediate state, corresponding to a highly ionized disk wind (log ξ > 6.1) with a velocity up to 0.06c. Meanwhile, quasi-simultaneous radio emissions have been detected before and after the X-rays, implying the co-existence of disk winds and jets in this system. Additionally, the observed wind in this source is potentially driven by magnetic forces. The absorption line disappeared in the soft state and a narrow emission line appeared at ~6.7 keV on top of the reflection component, which may be evidence for disk winds, but data with the higher spectral resolution are required to examine this.

astro-ph.HE↗

arXiv2020

Insight-HXMT observations of Swift J0243.6+6124: the evolution of RMS pulse fractions at super-Eddington luminosity

Based on Insight-HXMT data, we report on the pulse fraction evolution during the 2017-2018 outburst of the newly discovered first Galactic ultraluminous X-ray source (ULX) Swift J0243.6+6124. The pulse fractions of 19 observation pairs selected in the rising and fading phases with similar luminosity are investigated. The results show a general trend of the pulse fraction increasing with luminosity and energy at super-critical luminosity. However, the relative strength of the pulsation between each pair evolves strongly with luminosity. The pulse fraction in the rising phase is larger at luminosity below $7.71\times10^{38}$~erg~s$^{-1}$, but smaller at above. A transition luminosity is found to be energy independent. Such a phenomena is firstly confirmed by Insight-HXMT observations and we speculate it may have relation with the radiation pressure dominated accretion disk.

astro-ph.HE↗

arXiv2020

Physical origin of the nonphysical spin evolution of MAXI J1820+070

We report on the Insight-HXMT observations of the new black hole X-ray binary MAXI J1820+070 during its 2018 outburst. Detailed spectral analysis via the continuum fitting method shows an evolution of the inferred spin during its high soft sate. Moreover, the hardness ratio, the non-thermal luminosity and the reflection fraction also undergo an evolution, exactly coincident to the period when the inferred spin transition takes place. The unphysical evolution of the spin is attributed to the evolution of the inner disc, which is caused by the collapse of a hot corona due to condensation mechanism or may be related to the deceleration of a jet-like corona. The studies of the inner disc radius and the relation between the disc luminosity and the inner disc radius suggest that, only at a particular epoch, did the inner edge of the disc reach the innermost stable circular orbit and the spin measurement is reliable. We then constrain the spin of MAXI J1820+070 to be a*=0.2^{+0.2}_{-0.3}. Such a slowly spinning black hole possessing a strong jet suggests that its jet activity is driven mainly by the accretion disc rather than by the black hole spin.

astro-ph.HE↗

arXiv2020

X-ray reprocessing in accreting pulsar GX 301-2 observed with Insight-HXMT

We investigate the absorption and emission features in observations of GX 301-2 detected with Insight-HXMT/LE in 2017-2019. At different orbital phases, we found prominent Fe Kalpha, Kbeta and Ni Kalpha lines, as well as Compton shoulders and Fe K-shell absorption edges. These features are due to the X-ray reprocessing caused by the interaction between the radiation from the source and surrounding accretion material. According to the ratio of iron lines Kalpha and Kbeta, we infer the accretion material is in a low ionisation state. We find an orbital-dependent local absorption column density, which has a large value and strong variability around the periastron. We explain its variability as a result of inhomogeneities of the accretion environment and/or instabilities of accretion processes. In addition, the variable local column density is correlated with the equivalent width of the iron Kalpha lines throughout the orbit, which suggests that the accretion material near the neutron star is spherically distributed.

astro-ph.HE↗

arXiv2020

Insight-HXMT observations of a possible fast transition from jet to wind dominated state during a huge flare of GRS~1915+105

We present the analysis of the brightest flare that was recorded in the \emph{Insight}-HMXT data set, in a broad energy range (2$-$200 keV) from the microquasar GRS~1915+105 during an unusual low-luminosity state. This flare was detected by \emph{Insight}-HXMT among a series of flares during 2 June 2019 UTC 16:37:06 to 20:11:36, with a 2-200 keV luminosity of 3.4$-$7.27$\times10^{38}$ erg s$^{-1}$. Basing on the broad-band spectral analysis, we find that the flare spectrum shows different behaviors during bright and faint epochs. The spectrum of the flare can be fitted with a model dominated by a power-law component. Additional components show up in the bright epoch with a hard tail and in the faint epoch with an absorption line $\sim$ 6.78 keV. The reflection component of the latter is consistent with an inner disk radius $\sim$ 5 times larger than that of the former. These results on the giant flare during the "unusual" low-luminosity state of GRS~1915+105 may suggest that the source experiences a possible fast transition from a jet-dominated state to a wind-dominated state. We speculate that the evolving accretion disk and the large-scale magnetic field may play important roles in this peculiar huge flare.

astro-ph.HE↗

arXiv2020

Constraining the transient high-energy activity of FRB180916.J0158+65 with Insight-HXMT followup observations

A link between magnetars and fast radio burst (FRB) sources has finally been established. In this context, one of the open issues is whether/which sources of extra galactic FRBs exhibit X/gamma-ray outbursts and whether it is correlated with radio activity. We aim to constrain possible X/gamma-ray burst activity from one of the nearest extragalactic FRB sources currently known over a broad energy range, by looking for bursts over a range of timescales and energies that are compatible with being powerful flares from extragalactic magnetars. We followed up the as-yet nearest extragalactic FRB source at a mere 149 Mpc distance, the periodic repeater FRB180916.J0158+65, during the active phase on February 4-7, 2020, with the Insight-Hard X-ray Modulation Telescope (HXMT). Taking advantage of the combination of broad band, large effective area, and several independent detectors available, we searched for bursts over a set of timescales from 1 ms to 1.024 s with a sensitive algorithm, that had previously been characterised and optimised. Moreover, through simulations we studied the sensitivity of our technique in the released energy-duration phase space for a set of synthetic flares and assuming different energy spectra. We constrain the possible occurrence of flares in the 1-100 keV energy band to E<10^46 erg for durations <0.1 s over several tens of ks exposure. We can rule out the occurrence of giant flares similar to the ones that were observed in the few cases of Galactic magnetars. The absence of reported radio activity during our observations does not allow us to make any statements on the possible simultaneous high-energy emission.

astro-ph.HE↗