Papers made digestable
Our architecture simplifies the obstacle-perception
problem to that of place-dependent change detection. While we use the method with VT&R, it
can be generalized to suit arbitrary path-following applications.
Visual Teach and Repeat 3 (VT&R3), a generalization of stereo VT&R, achieves
long-term autonomous path-following using topometric mapping and localization
from a single rich sensor stream. In this paper, we improve the capabilities of
a LiDAR implementation of VT&R3 to reliably detect and avoid obstacles in
changing environments. Our architecture simplifies the obstacle-perception
problem to that of place-dependent change detection. We then extend the
behaviour of generic sample-based motion planners to better suit the
teach-and-repeat problem structure by introducing a new edge-cost metric paired
with a curvilinear planning space. The resulting planner generates naturally
smooth paths that avoid local obstacles while minimizing lateral path deviation
to best exploit prior terrain knowledge. While we use the method with VT&R, it
can be generalized to suit arbitrary path-following applications. Experimental
results from online run-time analysis, unit testing, and qualitative
experiments on a differential drive robot show the promise of the technique for
reliable long-term autonomous operation in complex unstructured environments.
Authors: Jordy Sehn, Yuchen Wu, Timothy D. Barfoot.
The statistical and design considerations that pertain to
dose optimization are discussed. The sample size savings range from 16.6% to 27.3%,
depending on the design and scenario, with a mean savings of 22.1%.
The traditional more-is-better dose selection paradigm, developed based on
cytotoxic chemotherapeutics, is often problematic When applied to the
development of novel molecularly targeted agents (e.g., kinase inhibitors,
monoclonal antibodies, and antibody-drug conjugates). The US Food and Drug
Administration (FDA) initiated Project Optimus to reform the dose optimization
and dose selection paradigm in oncology drug development and call for more
attention to benefit-risk consideration.
We systematically investigated the operating characteristics of the seamless
phase 2-3 design as a strategy for dose optimization, where in stage 1
(corresponding to phase 2) patients are randomized to multiple doses, with or
without a control; and in stage 2 (corresponding to phase 3) the efficacy of
the selected optimal dose is evaluated with a randomized concurrent control or
historical control. Depending on whether the concurrent control is included and
the type of endpoints used in stages 1 and 2, we describe four types of
seamless phase 2-3 dose-optimization designs, which are suitable for different
clinical settings. The statistical and design considerations that pertain to
dose optimization are discussed. Simulation shows that dose optimization phase
2-3 designs are able to control the familywise type I error rates and yield
appropriate statistical power with substantially smaller sample size than the
conventional approach. The sample size savings range from 16.6% to 27.3%,
depending on the design and scenario, with a mean savings of 22.1%. Due to the
interim dose selection, the phase 2-3 dose-optimization design is logistically
and operationally more challenging, and should be carefully planned and
implemented to ensure trial integrity.
Authors: Liyun Jiang, Ying Yuan.
We significantly improve performance using properties of the posterior
in our active learning scheme and for the definition of the GP prior. In
particular we account for the expected dynamical range of the posterior in
different dimensionalities. We test our model against a number of synthetic and
cosmological examples.
We present the GPry algorithm for fast Bayesian inference of general
(non-Gaussian) posteriors with a moderate number of parameters. GPry does not
need any pre-training, special hardware such as GPUs, and is intended as a
drop-in replacement for traditional Monte Carlo methods for Bayesian inference.
Our algorithm is based on generating a Gaussian Process surrogate model of the
log-posterior, aided by a Support Vector Machine classifier that excludes
extreme or non-finite values. An active learning scheme allows us to reduce the
number of required posterior evaluations by two orders of magnitude compared to
traditional Monte Carlo inference. Our algorithm allows for parallel
evaluations of the posterior at optimal locations, further reducing wall-clock
times. We significantly improve performance using properties of the posterior
in our active learning scheme and for the definition of the GP prior. In
particular we account for the expected dynamical range of the posterior in
different dimensionalities. We test our model against a number of synthetic and
cosmological examples. GPry outperforms traditional Monte Carlo methods when
the evaluation time of the likelihood (or the calculation of theoretical
observables) is of the order of seconds; for evaluation times of over a minute
it can perform inference in days that would take months using traditional
methods. GPry is distributed as an open source Python package (pip install
gpry) and can also be found at https://github.com/jonaselgammal/GPry.
Authors: Jonas El Gammal, Nils Schöneberg, Jesús Torrado, Christian Fidler.
We consider the fundamental scheduling problem of minimizing the sum of
weighted completion times on a single machine in the non-clairvoyant setting. However, to the best of our knowledge, this concept has never been considered
for the total completion time objective in the non-clairvoyant model. This implies
a performance guarantee of $(1+3\sqrt{3})\approx 6.197$ for the deterministic
algorithm and of $\approx 3.032$ for the randomized version.
We consider the fundamental scheduling problem of minimizing the sum of
weighted completion times on a single machine in the non-clairvoyant setting.
While no non-preemptive algorithm is constant competitive, Motwani, Phillips,
and Torng (SODA '93) proved that the simple preemptive round robin procedure is
$2$-competitive and that no better competitive ratio is possible, initiating a
long line of research focused on preemptive algorithms for generalized variants
of the problem. As an alternative model, Shmoys, Wein, and Williamson (FOCS
'91) introduced kill-and-restart schedules, where running jobs may be killed
and restarted from scratch later, and analyzed then for the makespan objective.
However, to the best of our knowledge, this concept has never been considered
for the total completion time objective in the non-clairvoyant model.
We contribute to both models: First we give for any $b > 1$ a tight analysis
for the natural $b$-scaling kill-and-restart strategy for scheduling jobs
without release dates, as well as for a randomized variant of it. This implies
a performance guarantee of $(1+3\sqrt{3})\approx 6.197$ for the deterministic
algorithm and of $\approx 3.032$ for the randomized version. Second, we show
that the preemptive Weighted Shortest Elapsed Time First (WSETF) rule is
$2$-competitive for jobs released in an online fashion over time, matching the
lower bound by Motwani et al. Using this result as well as the competitiveness
of round robin for multiple machines, we prove performance guarantees of
adaptions of the $b$-scaling algorithm to online release dates and unweighted
jobs on identical parallel machines.
Authors: Sven Jäger, Guillaume Sagnol, Daniel Schmidt genannt Waldschmidt, Philipp Warode.
Frozen pretrained models have become a viable alternative to the
pretraining-then-finetuning paradigm for transfer learning. With this work, we hope to
bring greater attention to this promising path of freezing pretrained image
models.
Frozen pretrained models have become a viable alternative to the
pretraining-then-finetuning paradigm for transfer learning. However, with
frozen models there are relatively few parameters available for adapting to
downstream tasks, which is problematic in computer vision where tasks vary
significantly in input/output format and the type of information that is of
value. In this paper, we present a study of frozen pretrained models when
applied to diverse and representative computer vision tasks, including object
detection, semantic segmentation and video action recognition. From this
empirical analysis, our work answers the questions of what pretraining task
fits best with this frozen setting, how to make the frozen setting more
flexible to various downstream tasks, and the effect of larger model sizes. We
additionally examine the upper bound of performance using a giant frozen
pretrained model with 3 billion parameters (SwinV2-G) and find that it reaches
competitive performance on a varied set of major benchmarks with only one
shared frozen base network: 60.0 box mAP and 52.2 mask mAP on COCO object
detection test-dev, 57.6 val mIoU on ADE20K semantic segmentation, and 81.7
top-1 accuracy on Kinetics-400 action recognition. With this work, we hope to
bring greater attention to this promising path of freezing pretrained image
models.
Authors: Yutong Lin, Ze Liu, Zheng Zhang, Han Hu, Nanning Zheng, Stephen Lin, Yue Cao.
While a large fraction of the stars are in multiple systems, our
understanding of the processes leading to the formation of these systems is
still inadequate. Given the large theoretical uncertainties, observation plays
a basic role.
While a large fraction of the stars are in multiple systems, our
understanding of the processes leading to the formation of these systems is
still inadequate. Given the large theoretical uncertainties, observation plays
a basic role. Here we discuss the contribution of high contrast imaging, and
more specifically of the SPHERE instrument at the ESO Very Large Telescope, in
this area. SPHERE nicely complements other techniques - in particular those
exploiting Gaia and ALMA - in detecting and characterising systems near the
peak of the distribution with separation and allows to capture snapshots of
binary formation within disks that are invaluable for the understanding of disk
fragmentation.
Authors: R. Gratton, S. Desidera, F. Marzari, M. Bonavita.
The construction of a meaningful hypergraph topology is the key to processing signals with high-order relationships that involve more than two entities. Finally, we conduct extensive experiments to evaluate the proposed framework on both synthetic and real world datasets. Experiments show that our proposed framework can efficiently infer meaningful hypergraph topologies from observed signals.
The construction of a meaningful hypergraph topology is the key to processing
signals with high-order relationships that involve more than two entities.
Learning the hypergraph structure from the observed signals to capture the
intrinsic relationships among the entities becomes crucial when a hypergraph
topology is not readily available in the datasets. There are two challenges
that lie at the heart of this problem: 1) how to handle the huge search space
of potential hyperedges, and 2) how to define meaningful criteria to measure
the relationship between the signals observed on nodes and the hypergraph
structure. In this paper, to address the first challenge, we adopt the
assumption that the ideal hypergraph structure can be derived from a learnable
graph structure that captures the pairwise relations within signals. Further,
we propose a hypergraph learning framework with a novel dual smoothness prior
that reveals a mapping between the observed node signals and the hypergraph
structure, whereby each hyperedge corresponds to a subgraph with both node
signal smoothness and edge signal smoothness in the learnable graph structure.
Finally, we conduct extensive experiments to evaluate the proposed framework on
both synthetic and real world datasets. Experiments show that our proposed
framework can efficiently infer meaningful hypergraph topologies from observed
signals.
Authors: Bohan Tang, Siheng Chen, Xiaowen Dong.
Due to time constraints and the high diversity of variants,
acous-tic measurements are more economical than vibration measurements. However, the acoustic data is affected by industrial disturbing noise. A real-world
dataset with typical faults and acoustic disturbances is recorded by an
acoustic array. Overall,
it is proposed to apply features extracted from a log-envelope spectrum
together with psychoacoustic features. The anomaly de-tection is done by using
the isolation forest or the more universal bagging random miner.
In the end-of-line test of geared motors, the evaluation of product qual-ity
is important. Due to time constraints and the high diversity of variants,
acous-tic measurements are more economical than vibration measurements.
However, the acoustic data is affected by industrial disturbing noise.
Therefore, the aim of this study is to investigate the robustness of features
used for anomaly detection in geared motor end-of-line testing. A real-world
dataset with typical faults and acoustic disturbances is recorded by an
acoustic array. This includes industrial noise from the production and
systematically produced disturbances, used to compare the robustness. Overall,
it is proposed to apply features extracted from a log-envelope spectrum
together with psychoacoustic features. The anomaly de-tection is done by using
the isolation forest or the more universal bagging random miner. Most
disturbances can be circumvented, while the use of a hammer or air pressure
often causes problems. In general, these results are important for condi-tion
monitoring tasks that are based on acoustic or vibration measurements.
Fur-thermore, a real-world problem description is presented to improve common
sig-nal processing and machine learning tasks.
Authors: Peter Wissbrock, David Pelkmann, Yvonne Richter.
We prove the equality of three conjectural formulas for the Brumer--Stark units.
We prove the equality of three conjectural formulas for the Brumer--Stark
units. The first formula has essentially been proven, so the present paper also
verifies the validity of the other two formulas.
Authors: Samit Dasgupta, Matthew H. Honnor.
Chiral magnets can host topological particles known as skyrmions which carry
an exactly quantised topological charge $Q=-1$. The mechanism behind this motion is similar to the one used
by a jellyfish when it swims through water. We show that the skyrmion's motion
is a universal phenomenon, arising in any magnetic system with translational
modes. For systems with small Gilbert
damping parameter $\alpha$, we identify two distinct physical mechanisms used
by the skyrmion to move.
Chiral magnets can host topological particles known as skyrmions which carry
an exactly quantised topological charge $Q=-1$. In the presence of an
oscillating magnetic field ${\bf B}_1(t)$, a single skyrmion embedded in a
ferromagnetic background will start to move with constant velocity ${\bf
v}_{\text{trans}}$. The mechanism behind this motion is similar to the one used
by a jellyfish when it swims through water. We show that the skyrmion's motion
is a universal phenomenon, arising in any magnetic system with translational
modes. By projecting the equation of motion onto the skyrmion's translational
modes and going to quadratic order in ${\bf B}_1(t)$ we obtain an analytical
expression for ${\bf v}_{\text{trans}}$ as a function of the system's linear
response. The linear response and consequently ${\bf v}_{\text{trans}}$ are
influenced by the skyrmion's internal modes and scattering states, as well as
by the ferromagnetic background's Kittel mode. The direction and speed of ${\bf
v}_{\text{trans}}$ can be controlled by changing the polarisation, frequency
and phase of the driving field ${\bf B}_1(t)$. For systems with small Gilbert
damping parameter $\alpha$, we identify two distinct physical mechanisms used
by the skyrmion to move. At low driving frequencies, the skyrmion's motion is
driven by friction, and $v_{\text{trans}}\sim\alpha$, whereas at higher
frequencies above the ferromagnetic gap the skyrmion moves by magnon emission
and $v_{\text{trans}}$ becomes independent of $\alpha$.
Authors: Nina del Ser, Vivek Lohani.
GPUs are essential to accelerating the latency-sensitive deep neural network (DNN) inference workloads in cloud datacenters. To fully utilize GPU resources, spatial sharing of GPUs among co-located DNN inference workloads becomes increasingly compelling. In this paper, we propose iGniter, an interference-aware GPU resource provisioning framework for cost-efficiently achieving predictable DNN inference in the cloud. We implement a prototype of iGniter based on the NVIDIA Triton inference server hosted on EC2 GPU instances.
GPUs are essential to accelerating the latency-sensitive deep neural network
(DNN) inference workloads in cloud datacenters. To fully utilize GPU resources,
spatial sharing of GPUs among co-located DNN inference workloads becomes
increasingly compelling. However, GPU sharing inevitably brings severe
performance interference among co-located inference workloads, as motivated by
an empirical measurement study of DNN inference on EC2 GPU instances. While
existing works on guaranteeing inference performance service level objectives
(SLOs) focus on either temporal sharing of GPUs or reactive GPU resource
scaling and inference migration techniques, how to proactively mitigate such
severe performance interference has received comparatively little attention. In
this paper, we propose iGniter, an interference-aware GPU resource provisioning
framework for cost-efficiently achieving predictable DNN inference in the
cloud. iGniter is comprised of two key components: (1) a lightweight DNN
inference performance model, which leverages the system and workload metrics
that are practically accessible to capture the performance interference; (2) A
cost-efficient GPU resource provisioning strategy that jointly optimizes the
GPU resource allocation and adaptive batching based on our inference
performance model, with the aim of achieving predictable performance of DNN
inference workloads. We implement a prototype of iGniter based on the NVIDIA
Triton inference server hosted on EC2 GPU instances. Extensive prototype
experiments on four representative DNN models and datasets demonstrate that
iGniter can guarantee the performance SLOs of DNN inference workloads with
practically acceptable runtime overhead, while saving the monetary cost by up
to 25% in comparison to the state-of-the-art GPU resource provisioning
strategies.
Authors: Fei Xu, Jianian Xu, Jiabin Chen, Li Chen, Ruitao Shang, Zhi Zhou, Fangming Liu.
Internet of Things (IoT) is a network of devices that communicate with each
other through the internet and provides intelligence to industry and people. These devices are running in potentially hostile environments, so the need for
security is critical. Trust Management aims to ensure the reliability of the
network by assigning a trust value in every node indicating its trust level.
Internet of Things (IoT) is a network of devices that communicate with each
other through the internet and provides intelligence to industry and people.
These devices are running in potentially hostile environments, so the need for
security is critical. Trust Management aims to ensure the reliability of the
network by assigning a trust value in every node indicating its trust level.
This paper presents an exhaustive survey of the current Trust Management
techniques for IoT, a classification based on the methods used in every work
and a discussion of the open challenges and future research directions.
Authors: Alyzia Maria Konsta, Alberto Lluch Lafuente, Nicola Dragoni.
The tetraquark system is solved by a correlated Gaussian method. Furthermore, based on the predicted tetraquark spectra we estimate their rearrangement decays in a quark-exchange model. We find that some of these couplings turn out to be sizeable.
In the framework of a nonrelativistic potential quark model, we investigate
the mass spectrum of the $1S$-wave charmed-strange tetraquark states of
$cn\bar{s}\bar{n}$ and $cs\bar{n}\bar{n}$ ($n=u$ or $d$) systems. The
tetraquark system is solved by a correlated Gaussian method. With the same
parameters fixed by the meson spectra, we obtained the mass spectra for the
$1S$-wave tetraquark states. Furthermore, based on the predicted tetraquark
spectra we estimate their rearrangement decays in a quark-exchange model. We
find that the resonances $X_0(2900)^0$ and $T^a_{c\bar{s}0}(2900)^{++/0}$
reported from LHCb may be assigned to be the lowest $1S$-wave tetraquark states
$\bar{T}_{cs0}^f(2818)$ and $T^{a}_{c\bar{s}0}(2828)$ classified in the quark
model, respectively. It also allows us to extract the couplings for the initial
tetraquark states to their nearby $S$-wave interaction channels. We find that
some of these couplings turn out to be sizeable. For $\bar{T}_{cs0}^f(2818)$
and $T^{a}_{c\bar{s}0}(2828)$ their couplings to $D^*\bar{K}^*$ and $D^*K^*$,
respectively, are found to be large. Following the picture of the wavefunction
renormalization for the near-threshold strong $S$-wave interactions, the
sizeable coupling strengths can be regarded as an indication of their dynamic
origins as candidates for hadronic molecules. Furthermore, our predictions
suggest that signals for the $1S$-wave charmed-strange tetraquark states can
also be searched in the other channels, such as $D^0K^+$, $D^+K^+$,
$D^{*+}K^-$, $D^{*+}K^+$, $D^{*0}K^+$, $D^0\bar{K}^{*0}$, $D_s^+\rho^0$, etc.
Authors: Feng-Xiao Liu, Ru-Hui Ni, Xian-Hui Zhong, Qiang Zhao.
We present a new description of the known large deviation function of the
classical symmetric simple exclusion process by exploiting its connection with
the quantum symmetric simple exclusion processes and using tools from free
probability. This latter result is obtained either by
developing a combinatorial approach for cumulants of products of random
variables or by borrowing techniques from Feynman graphs.
We present a new description of the known large deviation function of the
classical symmetric simple exclusion process by exploiting its connection with
the quantum symmetric simple exclusion processes and using tools from free
probability. This may seem paradoxal as free probability usually deals with non
commutative probability while the simple exclusion process belongs to the realm
of classical probability. On the way, we give a new formula for the free energy
-- alias the logarithm of the Laplace transform of the probability distribution
-- of correlated Bernoulli variables in terms of the set of their cumulants
with non-coinciding indices. This latter result is obtained either by
developing a combinatorial approach for cumulants of products of random
variables or by borrowing techniques from Feynman graphs.
Authors: Michel Bauer, Denis Bernard, Philippe Biane, Ludwig Hruza.
Through detailed finite-size scaling analysis, we study universality aspects of the non-equilibrium phase transition. Moreover, dynamic critical exponent of the local moves used in simulations is determined with high precision. Our numerical results are compatible with the previous ones on kinetic Ising models.
We perform extensive Monte Carlo simulations to investigate the dynamic phase
transition properties of the two-dimensional kinetic Ising model on the kagome
lattice in the presence of square-wave oscillating magnetic field. Through
detailed finite-size scaling analysis, we study universality aspects of the
non-equilibrium phase transition. Obtained critical exponents indicate that the
two-dimensional kagome-lattice kinetic Ising model belongs to the same
universality class with the corresponding Ising model in equilibrium. Moreover,
dynamic critical exponent of the local moves used in simulations is determined
with high precision. Our numerical results are compatible with the previous
ones on kinetic Ising models.
Authors: Zeynep Demir Vatansever.