Search

We study the axion-photon conversion process in pulsar binary systems with high Shapiro time delay. In these binary systems, the orbital plane is nearly parallel to the line of sight. When the companion star is positioned between the pulsar and the Earth, the pulsar radiation beam will pass through the companion's magnetic field, potentially leading to axion-photon conversion that affects the intensity and linear polarization of the photons. The advantage of using such systems for axion or axion-like particle (ALP) research is that the intrinsic intensity and polarization state of the source photons can be well determined. This work analyzes the axion-photon conversion and the magnetic field along the photon path, establishing the corresponding model and solving it numerically. We choose PSR J1614-2230 and PSR J1910-5959A for our study. By assuming the companion's magnetic field, magnetic axis angle, and detector sensitivity, we discuss the feasibility of using high Shapiro delay binary systems to constrain the axion/ALP parameter space. We find that the upper limits derived from these two sources are weaker then existing constraints. However, considering the higher reliability, the method proposed in this work is still valuable. As more suitable samples and richer observational results are discovered, further constraints on the axion parameters could be strengthened.

We investigate the scattering halos resulting from collisions between discrete momentum components in the time-of-flight expansion of interaction-tunable $^6\rm Li_2$ molecular Bose-Einstein condensates. A key highlight of this study is the observation of the influence of interactions on the collisional scattering process. We measure the production of scattering halos at different interaction levels by varying the number of particles and the scattering length, and quantitatively assess the applicability of perturbation theory. To delve into a general theory of scattering halos, we introduce a scattering factor and obtain a universal relation between it and the halo ratio. Furthermore, we simulate the formation of scattering halos under non-perturbative conditions and analyze the discrepancies between simulation results and experiments through a return pulse experiment. This study enhances our understanding of the physical mechanisms underlying scattering processes in many-body systems and provides new perspectives for further theoretical research.

The scalar meson $f_{0}(980)$ has long posed a perplexing puzzle within the realm of light hadron physics. Conventionally, its mass and width in normal decay processes have been estimated as $M=990\pm20$~MeV/$c^2$ and $\Gamma=40-100$~MeV, respectively. Theoretical explanations regarding the internal structure of $f_{0}(980)$ range from it being a conventional quark-antiquark meson to a tetraquark state, a $K\overline{K}$ molecule, or even a quark-antiquark gluon hybrid. However, a definitive consensus has remained elusive over a considerable duration. Recent observations by the BESIII experiment have unveiled anomalously narrow widths of $f_{0}(980)$ in five independent isospin-symmetry-breaking decay channels. Harnessing these experimental findings, we performed a simultaneous fit to the $\pi\pi$ invariant mass distributions, resulting in a refined determination of the mass and width in isospin-symmetry-breaking decays as $M=990.0\pm0.4(\text{stat})\pm0.1(\text{syst})$~MeV/$c^2$ and $\Gamma=11.4\pm1.1(\text{stat})\pm0.9(\text{syst})$~MeV, respectively. Here, the first errors are statistical and the second are systematic. Furthermore, by employing the parameterized Flatt\'{e} formula to fit the same $\pi\pi$ invariant mass distributions, we ascertained the values of the two coupling constants, $g_{f\pi\pi}$ and $g_{fK\overline{K}}$, as $g_{f\pi\pi}=0.46\pm0.03$ and $g_{fK\overline{K}}=1.24\pm0.32$, respectively. Based on the joint confidence regions of $g_{f\pi\pi}$ and $g_{fK\overline{K}}$, we draw the conclusion that the experimental data exhibit a propensity to favor the $K\overline{K}$ molecule model and the quark-antiquark ($q\bar{q}$) model, while offering relatively less support for the tetraquarks ($q^{2}\bar{q}^{2}$) model and the quark-antiquark gluon ($q\bar{q}g$) hybrid model., Comment: 15 pages, 3 figures

The excited bands in optical lattices manifest an important tool for studying quantum simulation and many-body physics, making it crucial to measure high-band scattering dynamics under strong interactions. This work investigates both experimentally and theoretically the collisional scattering of $^{6}\rm Li_2$ molecular Bose-Einstein condensate in the $D$ band of a one-dimensional optical lattice, with interaction strength directly tunable via magnetic Feshbach resonance. We find a clear dependence of the $D$-band lifetimes on the interaction strength within the strongly interacting regime, which arises from the fact that the scattering cross-section is proportional to the square of the scattering length. The maximum lifetime versus lattice depth is measured to reveal the effects of interactions. We also investigate the scattering channels of $D$-band molecules under different interaction levels and develop a reliable two-body scattering rate equation. This work provides insight into the interplay between interaction and the collisional scattering of high-band bosons in optical lattices, paving the way for research into strong correlation effects in high-band lattice systems.

Stable diffusion plays a crucial role in generating high-quality images. However, image generation is time-consuming and memory-intensive. To address this, stable-diffusion.cpp (Sdcpp) emerges as an efficient inference framework to accelerate the diffusion models. Although it is lightweight, the current implementation of ggml_conv_2d operator in Sdcpp is suboptimal, exhibiting both high inference latency and massive memory usage. To address this, in this work, we present an optimized version of Sdcpp leveraging the Winograd algorithm to accelerate 2D convolution operations, which is the primary bottleneck in the pipeline. By analyzing both dependent and independent computation graphs, we exploit the device's locality and parallelism to achieve substantial performance improvements. Our framework delivers correct end-to-end results across various stable diffusion models, including SDv1.4, v1.5, v2.1, SDXL, and SDXL-Turbo. Our evaluation results demonstrate a speedup up to 2.76x for individual convolutional layers and an inference speedup up to 4.79x for the overall image generation process, compared with the original Sdcpp on M1 pro. Homepage: https://github.com/SealAILab/stable-diffusion-cpp

Tidal disruption events (TDEs), where stars are captured or tidally disrupted by supermassive black holes, are potential sources of high-energy neutrinos. We report the discovery of a potential neutrino flare that is spatially and temporally associated with X-ray emission from TDE ATLAS17jrp. The best-fit spectrum of the neutrino flare follows a power-law with an index of $\rm{\gamma=2.7\pm0.4}$ and a flux normalization of $\rm{\Phi_0 =1.7^{+6.3}_{-1.5}\times 10^{-18}\;GeV^{-1} cm^{-2} s^{-1}}$ at 100 TeV based on an analysis of 10-year track data from IceCube, and the flare duration is 61 days. We calculate that the probability of this association occurring by chance is $0.17\%$. Therefore, ATLAS17jrp is the second TDE (not including candidates) associated with high-energy neutrinos, following TDE AT2019dsg associated with an IceCube neutrino alert. This association can be attributed to the interaction of X-ray photons produced by the hot corona with high-energy particles accelerated by disk winds or outflows, resulting in the production of neutrinos., Comment: 8 pages, 2 figures, submitted to ApJL

In this paper, we propose MCUBERT to enable language models like BERT on tiny microcontroller units (MCUs) through network and scheduling co-optimization. We observe the embedding table contributes to the major storage bottleneck for tiny BERT models. Hence, at the network level, we propose an MCU-aware two-stage neural architecture search algorithm based on clustered low-rank approximation for embedding compression. To reduce the inference memory requirements, we further propose a novel fine-grained MCU-friendly scheduling strategy. Through careful computation tiling and re-ordering as well as kernel design, we drastically increase the input sequence lengths supported on MCUs without any latency or accuracy penalty. MCUBERT reduces the parameter size of BERT-tiny and BERT-mini by 5.7$\times$ and 3.0$\times$ and the execution memory by 3.5$\times$ and 4.3$\times$, respectively. MCUBERT also achieves 1.5$\times$ latency reduction. For the first time, MCUBERT enables lightweight BERT models on commodity MCUs and processing more than 512 tokens with less than 256KB of memory., Comment: ICCAD 2024

Memory-based video object segmentation methods model multiple objects over long temporal-spatial spans by establishing memory bank, which achieve the remarkable performance. However, they struggle to overcome the false matching and are prone to lose critical information, resulting in confusion among different objects. In this paper, we propose an effective approach which jointly improving the matching and decoding stages to alleviate the false matching issue.For the memory matching stage, we present a cost aware mechanism that suppresses the slight errors for short-term memory and a shunted cross-scale matching for long-term memory which establish a wide filed matching spaces for various object scales. For the readout decoding stage, we implement a compensatory mechanism aims at recovering the essential information where missing at the matching stage. Our approach achieves the outstanding performance in several popular benchmarks (i.e., DAVIS 2016&2017 Val (92.4%&88.1%), and DAVIS 2017 Test (83.9%)), and achieves 84.8%&84.6% on YouTubeVOS 2018&2019 Val., Comment: Accepted to ICPR2024

Primordial black holes (PBHs) are hypothetical objects formed at the early epoch of the universe, which could be a type of dark matter (DM) candidate without the need for new particles. The abundance of PBH DM has been constrained strictly by many observations.In this work, with the radio observations of Fornax and Segue I, we constrain the abundance of PBH in dwarf spheroidal galaxies through the synchrotron self-Compton (SSC) effect of Hawking radiation electrons. By selecting optimal sources, we obtain the constraints on the fraction of PBH DM down to $\sim10^{-3}$ for Segue I and $\sim10^{-5}$ for Fornax at asteroidal mass. We also predict that, with 100 hours of future observation by the Square Kilometer Array, the SSC approach could place constraints comparable to the current strictest results for PBHs of $<5\times10^{15}\,{\rm g}$. Better projected constraints can be obtained by including the inverse Compton scattering on cosmic microwave background photons., Comment: 9 pages, 6 figures, accepted for publication in PRD

As the largest gravitationally bound objects in the Universe, galaxy clusters have provided the first piece of evidence for the presence of dark matter and may be suitable targets for indirect dark matter searches. Among various signals, the GeV-TeV $\gamma$-ray line has been taken as the smoking-gun signal of the dark matter annihilation/decay since no known astrophysical/physical process(es) could generate such a peculiar spectrum. With 15.5 years of Fermi-LAT P8R3 publicly available data, we search for the $\gamma$-ray line emission in the directions of a group of 13 nearby massive galaxy clusters with an unbinned likelihood analysis. A $\gamma$-ray line signal at $\sim 43.2$ GeV has a net TS value of $\approx 30$ if we only take into account the data in the directions of Virgo, Fornax and Ophiuchus clusters, three massive clusters with the highest J-factors expected to generate the dark matter annihilation signal. The signal still presents when the data of other 10 nearby massive clusters have also been included, though the TS value decreases to $\approx 21$ likely because of their lower signal-to-noise ratios. The absence of this signal in the inner Galaxy disfavors both the instrumental effect and the canonical dark matter annihilation interpretation, and a more sophisticated dark matter model or very peculiar astrophysical scenario might be needed. This $\gamma$-ray line signal, if intrinsic, could be unambiguously verified by the Very Large Area $\gamma$-ray Space Telescope in its first two years of performance., Comment: 9 pages, 7 figures, 1 table. Comments are welcome!

Superluminous supernovae (SLSNe) are a new class of transients with luminosities $\sim10 -100$ times larger than the usual core-collapse supernovae (SNe). Their origin is still unclear and one widely discussed scenario involves a millisecond magnetar central engine. The GeV-TeV emission of SLSNe has been predicted in the literature but has not been convincingly detected yet. Here we report the search for the $\gamma$-ray emission in the direction of SN 2017egm, one of the closest SLSNe detected so far, with the 15-year {\it Fermi}-LAT Pass 8 data. There is a transient $\gamma$-ray source appearing about 2 months after this event and lasting a few months. Both the peak time and the luminosity of the GeV emission are consistent with the magnetar model prediction, suggesting that such a GeV transient is the high-energy counterpart of SN 2017egm and the central engine of this SLSNe is a young magnetar.

The emergence of the large language model~(LLM) poses an exponential growth of demand for computation throughput, memory capacity, and communication bandwidth. Such a demand growth has significantly surpassed the improvement of corresponding chip designs. With the advancement of fabrication and integration technologies, designers have been developing Wafer-Scale Chips~(WSCs) to scale up and exploit the limits of computation density, memory capacity, and communication bandwidth at the level of a single chip. Existing solutions have demonstrated the significant advantages of WSCs over traditional designs, showing potential to effectively support LLM workloads. Despite the benefits, exploring the early-stage design space of WSCs for LLMs is a crucial yet challenging task due to the enormous and complicated design space, time-consuming evaluation methods, and inefficient exploration strategies. To address these challenges, we propose Theseus, an efficient WSC design space exploration framework for LLMs. We construct the design space of WSCs with various constraints considering the unique characteristics of WSCs. We propose efficient evaluation methodologies for large-scale NoC-based WSCs and introduce multi-fidelity Bayesian optimization to efficiently explore the design space. Evaluation results demonstrate the efficiency of Theseus that the searched Pareto optimal results outperform GPU cluster and existing WSC designs by up to 62.8\%/73.7\% in performance and 38.6\%/42.4\% in power consumption for LLM training, while improving up to 23.2$\times$ and 15.7$\times$ for the performance and power of inference tasks. Furthermore, we conduct case studies to address the design tradeoffs in WSCs and provide insights to facilitate WSC designs for LLMs.

Current unsupervised anomaly detection approaches perform well on public datasets but struggle with specific anomaly types due to the domain gap between pre-trained feature extractors and target-specific domains. To tackle this issue, this paper presents a two-stage training strategy, called \textbf{ToCoAD}. In the first stage, a discriminative network is trained by using synthetic anomalies in a self-supervised learning manner. This network is then utilized in the second stage to provide a negative feature guide, aiding in the training of the feature extractor through bootstrap contrastive learning. This approach enables the model to progressively learn the distribution of anomalies specific to industrial datasets, effectively enhancing its generalizability to various types of anomalies. Extensive experiments are conducted to demonstrate the effectiveness of our proposed two-stage training strategy, and our model produces competitive performance, achieving pixel-level AUROC scores of 98.21\%, 98.43\% and 97.70\% on MVTec AD, VisA and BTAD respectively., Comment: 11 pages, 7 figures

IoT devices based on microcontroller units (MCU) provide ultra-low power consumption and ubiquitous computation for near-sensor deep learning models (DNN). However, the memory of MCU is usually 2-3 orders of magnitude smaller than mobile devices, which makes it challenging to map DNNs onto MCUs. Previous work separates memory management and kernel implementation for MCU and relies on coarse-grained memory management techniques such as inplace update to reduce memory consumption. In this paper, we propose to coordinate memory management and kernel optimization for DNN inference on MCUs to enable fine-grained memory management. The key idea is to virtualize the limited memory of MCU as a large memory pool. Each kernel divides the memory pool into kernel-specific segments and handles segment load and store while computing DNN layers. Memory consumption can be reduced because using the fine-grained segment-level memory control, we can overlap the memory footprint of different tensors without the need to materialize them at the same time. Following this idea, we implement \ours{} for DNN inference on MCU. Evaluation for single layers on ARM Cortex-M4 and Cortex-M7 processors shows that \ours{} can reduce from $12.0\%$ to $49.5\%$ RAM usage and from $20.6\%$ to $53.0\%$ energy consumption compared to state-of-the-art work. For full DNN evaluation, \ours{} can reduce the memory bottleneck by $61.5\%$, enabling more models to be deployed on low-end MCUs.

Given that gamma rays with energies larger than TeV are severely absorbed by background radiation fields, for many extragalactic sources, the GeV-TeV gamma-ray observations are the messengers that are closest in energy to the TeV-PeV neutrinos observed by IceCube. Investigating whether there is a correlation between the gamma-ray and neutrino observations can help us identify high-energy neutrino sources and determine which sources are the main contributors to the all-sky diffuse neutrino flux of IceCube. In previous work, we have already studied the possible gamma-neutrino correlations by analyzing 10 years of IceCube muon-track data. In this work, we further investigate such correlations by employing the IceCube p-value sky map of the scan for point sources. We examine the spatial associations of hotspots in the neutrino sky map with various gamma-ray source samples: the third Fermi-LAT catalog of high-energy sources (3FHL), LAT 14-year source catalog (4FGL), the fourth catalog of active galactic nuclei (4LAC) and subsets of these samples. Among all the samples, the 3FHL sample shows a possible correlation with the neutrino hotspots, with a pre-trial p-value of $1.1\times10^{-4}$ ($\sim 3.9\,\sigma$). However, this is found to be caused by three already known neutrino sources/source candidates: NGC 1068, TXS 0506+056, and PKS 1424+240. In order to validate our analysis procedure and to test the robustness of the previously claimed correlation between the 5BZCAT blazars and neutrino hotspots, we also consider 5BZCAT blazars in our correlation study. We find that the way in which mock sources are generated in the simulation used to derive the chance coincidence probability may have a large impact on the claimed correlation., Comment: Prepared for submission, comments are welcome

Within the Electronic Design Automation (EDA) domain, AI-driven solutions have emerged as formidable tools, yet they typically augment rather than redefine existing methodologies. These solutions often repurpose deep learning models from other domains, such as vision, text, and graph analytics, applying them to circuit design without tailoring to the unique complexities of electronic circuits. Such an AI4EDA approach falls short of achieving a holistic design synthesis and understanding, overlooking the intricate interplay of electrical, logical, and physical facets of circuit data. This paper argues for a paradigm shift from AI4EDA towards AI-native EDA, integrating AI at the core of the design process. Pivotal to this vision is the development of a multimodal circuit representation learning technique, poised to provide a comprehensive understanding by harmonizing and extracting insights from varied data sources, such as functional specifications, RTL designs, circuit netlists, and physical layouts. We champion the creation of large circuit models (LCMs) that are inherently multimodal, crafted to decode and express the rich semantics and structures of circuit data, thus fostering more resilient, efficient, and inventive design methodologies. Embracing this AI-native philosophy, we foresee a trajectory that transcends the current innovation plateau in EDA, igniting a profound shift-left in electronic design methodology. The envisioned advancements herald not just an evolution of existing EDA tools but a revolution, giving rise to novel instruments of design tools that promise to radically enhance design productivity and inaugurate a new epoch where the optimization of circuit performance, power, and area (PPA) is achieved not incrementally, but through leaps that redefine the benchmarks of electronic systems' capabilities., Comment: The authors are ordered alphabetically. Contact: qxu@cse[dot]cuhk[dot]edu[dot]hk, gluo@pku[dot]edu[dot]cn, yuan.mingxuan@huawei[dot]com

Talbot interferometer, as a periodic reproduction of momentum distribution in the time domain, finds significant applications in multiple research. The inter-particle interactions during the diffraction and interference process introduce numerous many-body physics problems, leading to unconventional interference characteristics. This work investigates both experimentally and theoretically the influence of interaction in a Talbot interferometer with a $^{6}\rm Li_2$ molecular Bose-Einstein condensate in a one-dimensional optical lattice, with interaction strength directly tunable via magnetic Feshbach resonance. A clear dependence of the period and amplitude of signal revivals on the interaction strength can be observed. While interactions increase the decay rate of the signal and advance the revivals, we find that over a wide range of interactions, the Talbot interferometer remains highly effective over a certain evolutionary timescale, including the case of fractional Talbot interference. This work provides insight into the interplay between interaction and the coherence properties of a temporal Talbot interference in optical lattices, paving the way for research into quantum interference in strongly interacting systems.

Recently, Transformers have been introduced into the field of acoustics recognition. They are pre-trained on large-scale datasets using methods such as supervised learning and semi-supervised learning, demonstrating robust generality--It fine-tunes easily to downstream tasks and shows more robust performance. However, the predominant fine-tuning method currently used is still full fine-tuning, which involves updating all parameters during training. This not only incurs significant memory usage and time costs but also compromises the model's generality. Other fine-tuning methods either struggle to address this issue or fail to achieve matching performance. Therefore, we conducted a comprehensive analysis of existing fine-tuning methods and proposed an efficient fine-tuning approach based on Adapter tuning, namely AAT. The core idea is to freeze the audio Transformer model and insert extra learnable Adapters, efficiently acquiring downstream task knowledge without compromising the model's original generality. Extensive experiments have shown that our method achieves performance comparable to or even superior to full fine-tuning while optimizing only 7.118% of the parameters. It also demonstrates superiority over other fine-tuning methods., Comment: Preprint version for ICASSP 2024, Korea

Dwarf irregular galaxies (dIrrs), as rotationally-supported systems, have more reliable J-factor measurements than dwarf spheroidal galaxies and have received attention as targets for dark matter detection in recent years. In this paper, we use 10 years of IceCube muon-track data and an unbinned maximum-likelihood-ratio method to search for neutrino signals beyond the background from the directions of 7 dIrrs, aiming to detect neutrinos produced by heavy annihilation dark matter. We do not detect any significant signal. Based on such null results, we calculate the upper limits on the velocity-averaged annihilation cross section for 1 TeV - 10 PeV dark matter. Our limits, although weaker than the strictest constraints in the literature in this mass range, are also a good complement to the existing results considering the more reliable J-factor measurements of dIrrs., Comment: 13 pages, 4 figures, accepted for publication in RAA

The physical origin of fast radio bursts (FRBs) is still unclear. However, young magnetars associated with short-duration gamma-ray bursts (SGRBs) have been thought to be possible central engines for some FRBs. In this paper, we perform a systematic search for SGRBs that are associated with FRBs in a sample including 623 FRBs (601 one-off bursts and 22 repeaters) and 168 SGRBs with precise localizations. We find that FRB 190309A is spatially associated with GRB 060502B, with a chance probability of 0.05 when temporal and redshift information is taken into account. Considering the high chance probability (the statistical significance is < 3{\sigma}), we examine other observational properties such as the host galaxy, the dispersion measure, and the energy budget of the central engine to check the possibility of their association. Although the available observational information is insufficient to determine whether they are physically associated, it does not rule out such a possibility. As the only pair of FRB and GRB that are spatially associated, it remains an interesting case worthy of further attention, Comment: 19 pages, 4 figures, accepted to ApJ

Electrons and positrons produced in dark matter annihilation can generate secondary emission through synchrotron and IC processes, and such secondary emission provides a possible means to detect DM particles with masses beyond the detector's energy band. The secondary emission of heavy dark matter (HDM) particles in the TeV-PeV mass range lies within the Fermi-LAT energy band. In this paper, we utilize the Fermi-LAT observations of dwarf spheroidal (dSph) galaxies to search for annihilation signals of HDM particles. We consider the propagation of $e^+/e^-$ produced by DM annihilation within the dSphs, derive the electron spectrum of the equilibrium state by solving the propagation equation, and then compute the gamma-ray signals produced by the $e^+/e^-$ population through the IC and synchrotron processes. Considering the spatial diffusion of electrons, the dSphs are modeled as extended sources in the analysis of Fermi-LAT data according to the expected spatial intensity distribution of the gamma rays. We do not detect any significant HDM signal. By assuming a magnetic field strength of $B=1\,{\rm \mu G}$ and a diffusion coefficient of $D_0 = 3\times10^{28}\,{\rm cm^{2}s^{-1}}$ of the dSphs, we place limits on the annihilation cross section for HDM particles. Our results are weaker than the previous limits given by the VERITAS and IceCube observations of dSphs, but extend the existing limits to higher DM masses. As a complement, we also search for the prompt $\gamma$-rays produced by DM annihilation and give limits on the cross section in the 10-$10^5$ GeV mass range. Consequently, in this paper we obtain the upper limits on the DM annihilation cross section for a very wide mass range from 10 GeV to 100 PeV in a unified framework of the Fermi-LAT data analysis., Comment: 11 pages, 9 figures; add the KN effect for IC and model the secondary emission around dSphs as extended

Weakly interacting massive particles, as a major candidate of dark matter (DM), may directly annihilate or decay into high-energy photons, producing monochromatic spectral lines in the gamma-ray band. These spectral lines, if detected, are smoking-gun signatures for the existence of new physics. Using the 5 years of DAMPE and 13 years of Fermi-LAT data, we search for line-like signals in the energy range of 3 GeV to 1 TeV from the Galactic halo. Different regions of interest are considered to accommodate different DM density profiles. We do not find any significant line structure, and the previously reported line-like feature at $\sim$133 GeV is also not detected in our analysis. Adopting a local DM density of $\rho_{\rm local}=0.4\,{\rm GeV\,cm^{-3}}$, we derive 95% confidence level constraints on the velocity-averaged cross-section of $\langle{\sigma v}\rangle_{\gamma\gamma} \lesssim 4 \times 10^{-28}\,{\rm cm^{3}\,s^{-1}}$ and the decay lifetime of $\tau_{\gamma\nu} \gtrsim 5 \times 10^{29}\,{\rm s}$ at 100 GeV, achieving the strongest constraints to date for the line energies of 6-660 GeV. The improvement stems from the longer Fermi-LAT data set used and the inclusion of DAMPE data in the analysis. The simultaneous use of two independent data sets could also reduce the systematic uncertainty of the search., Comment: 14 pages, 10 figures. Accepted for publication in PRD

Time-resolved illumination provides rich spatio-temporal information for applications such as accurate depth sensing or hidden geometry reconstruction, becoming a useful asset for prototyping and as input for data-driven approaches. However, time-resolved illumination measurements are high-dimensional and have a low signal-to-noise ratio, hampering their applicability in real scenarios. We propose a novel method to compactly represent time-resolved illumination using mixtures of exponentially-modified Gaussians that are robust to noise and preserve structural information. Our method yields representations two orders of magnitude smaller than discretized data, providing consistent results in applications such as hidden scene reconstruction and depth estimation, and quantitative improvements over previous approaches.

Memory-aware network scheduling is becoming increasingly important for deep neural network (DNN) inference on resource-constrained devices. However, due to the complex cell-level and network-level topologies, memory-aware scheduling becomes very challenging. While previous algorithms all suffer from poor scalability, in this paper, we propose an efficient memory-aware scheduling framework based on iterative computation graph optimization. Our framework features an iterative graph fusion algorithm that simplifies the computation graph while preserving the scheduling optimality. We further propose an integer linear programming formulation together with topology-aware variable pruning to schedule the simplified graph efficiently. We evaluate our method against prior-art algorithms on different networks and demonstrate that our method outperforms existing techniques in all the benchmarks, reducing the peak memory footprint by 13.4%, and achieving better scalability for networks with complex network-level topologies.