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.
Both results are tight up to the value of $C$ and they answer two recent
questions posed by Kam\v{c}ev and M\"{u}yesser.
We show that there is a constant $C$ such that for every $\varepsilon>0$ any
$2$-coloured $K_n$ with minimum degree at least $n/4+\varepsilon n$ in both
colours contains a complete subgraph on $2t$ vertices where one colour class
forms a $K_{t,t}$, provided that $n\geq \varepsilon^{-Ct}$. Also, we prove that
if $K_n$ is $2$-coloured with minimum degree at least $\varepsilon n$ in both
colours then it must contain one of two natural colourings of a complete graph.
Both results are tight up to the value of $C$ and they answer two recent
questions posed by Kam\v{c}ev and M\"{u}yesser.
Authors: António Girão, David Munhá Correia.
Mass transfer stability is an essential issue in binary evolution. Ge et al. We would investigate the influence of mass transfer stability on the formation and properties of DWD populations. While the polytropic model overpredicts space density of DWDs by a factor of about $2-3$.
Mass transfer stability is an essential issue in binary evolution. Ge et al.
studied critical mass ratios for dynamically stable mass transfer by
establishing adiabatic mass loss model and found that the donor stars on the
giant branches tend to be more stable than that based on the composite
polytropic stellar model. We would investigate the influence of mass transfer
stability on the formation and properties of DWD populations. We performed a
series of binary population synthesis, where the critical mass ratios from
adiabatic mass loss model (Ge's model) and that from the composite polytropic
model are adopted, respectively. For Ge's model, most of the DWDs are produced
from the stable non-conservative Roche lobe overflow plus common envelope (CE)
ejection channel (RL+CE channel) regardless of the CE ejection efficiency
$\alpha_{CE}$. While the results of the polytropic model strongly depend on the
adopted value of $\alpha_{ CE}$. We find DWDs produced from the RL+CE channel
have comparable WD masses and the mass ratio distribution peaks at around 1.
Based on the magnitude-limited sample of DWDs, the space densities for the
detectable DWDs and those with extremely low-mass WD (ELM WD) companions in
Ge's model is $1347$ and $473 kpc^{-3}$, respectively, close to observations.
While the polytropic model overpredicts space density of DWDs by a factor of
about $2-3$. We also find that the results of DWD merger rate distribution in
Ge's model reproduce the observations better than that of the polytropic model,
and the merger rate of DWDs with ELM WD companions in the Galaxy is about
$1.8\times 10^{-3} yr^{-1}$ in Ge's model, which is comparable to the
observation estimation. We confirm that the mass transfer stability plays
important roles in the formation and properties of DWD populations, and then in
the progenitors of SNe Ia and detectable GW sources.
Authors: Zhenwei LI, Xuefei Chen, Hongwei Ge, Hai-Liang Chen, Zhanwen Han.
These constraints
are then enforced onto the RL updates in an effort to enhance the learning
method with a probabilistic safety mechanism. We illustrate the results of our method in a numerical example on
the control of a quadrotor drone in a safety-critical environment.
In this work, we propose a method to encourage safety in a Model Predictive
Control (MPC)-based Reinforcement Learning (RL) agent via Gaussian Process (GP)
regression. This framework consists of 1) a parametric MPC scheme that is
employed as model-based controller with approximate knowledge on the real
system's dynamics, 2) an episodic RL algorithm tasked with adjusting the MPC
parametrization in order to increase its performance, and lastly, 3) GP
regressors used to estimate, directly from data, constraints on the MPC
parameters capable of predicting, up to some probability, whether the
parametrization is likely to yield a safe or unsafe policy. These constraints
are then enforced onto the RL updates in an effort to enhance the learning
method with a probabilistic safety mechanism. Compared to other recent
publications combining safe RL with MPC, our method does not require further
assumptions on, e.g., the prediction model in order to retain computational
tractability. We illustrate the results of our method in a numerical example on
the control of a quadrotor drone in a safety-critical environment.
Authors: Filippo Airaldi, Bart De Schutter, Azita Dabiri.
CTCA is ideal for geometry reconstruction to create virtual models of coronary arteries. To our knowledge there is no public dataset that includes centrelines and segmentation of the full coronary tree. Manual voxel-wise segmentations by three experts were combined using majority voting to generate the final annotations.
Computed Tomography Coronary Angiography (CTCA) is a non-invasive method to
evaluate coronary artery anatomy and disease. CTCA is ideal for geometry
reconstruction to create virtual models of coronary arteries. To our knowledge
there is no public dataset that includes centrelines and segmentation of the
full coronary tree.
We provide anonymized CTCA images, voxel-wise annotations and associated data
in the form of centrelines, calcification scores and meshes of the coronary
lumen in 20 normal and 20 diseased cases. Images were obtained along with
patient information with informed, written consent as part of Coronary Atlas
(https://www.coronaryatlas.org/). Cases were classified as normal (zero calcium
score with no signs of stenosis) or diseased (confirmed coronary artery
disease). Manual voxel-wise segmentations by three experts were combined using
majority voting to generate the final annotations.
Provided data can be used for a variety of research purposes, such as 3D
printing patient-specific models, development and validation of segmentation
algorithms, education and training of medical personnel and in-silico analyses
such as testing of medical devices.
Authors: Ramtin Gharleghi, Dona Adikari, Katy Ellenberger, Mark Webster, Chris Ellis, Arcot Sowmya, Sze-Yuan Ooi, Susann Beier.
Graph auto-encoders are widely used to construct graph representations in
Euclidean vector spaces. So why are we still using
nonlinear graph auto-encoders? Our experiments
show that the linear encoder can outperform the nonlinear encoder when using
feature information.
Graph auto-encoders are widely used to construct graph representations in
Euclidean vector spaces. However, it has already been pointed out empirically
that linear models on many tasks can outperform graph auto-encoders. In our
work, we prove that the solution space induced by graph auto-encoders is a
subset of the solution space of a linear map. This demonstrates that linear
embedding models have at least the representational power of graph
auto-encoders based on graph convolutional networks. So why are we still using
nonlinear graph auto-encoders? One reason could be that actively restricting
the linear solution space might introduce an inductive bias that helps improve
learning and generalization. While many researchers believe that the
nonlinearity of the encoder is the critical ingredient towards this end, we
instead identify the node features of the graph as a more powerful inductive
bias. We give theoretical insights by introducing a corresponding bias in a
linear model and analyzing the change in the solution space. Our experiments
show that the linear encoder can outperform the nonlinear encoder when using
feature information.
Authors: Solveig Klepper, Ulrike von Luxburg.
The total symmetry of the Dirac equation is the symmetry $ SO(3) \otimes SU(2) $. The generator of the $ SO(3) $ symmetry group is given by the total momentum operator, and the generator of $ SU(2) $ group is given by the rotation of the vector-states in the spinor space, determined by the Dirac, Johnson-Lippmann, and the new spinor invariants.
It is shown that the Dirac equation with the Coulomb potential can be solved
using the algebra of the three spinor invariants of the Dirac equation without
the involvement of the methods of supersymmetric quantum mechanics. The Dirac
Hamiltonian is invariant with respect to the rotation transformation, which
indicates the dynamical (hidden) symmetry $ SU(2) $ of the Dirac equation. The
total symmetry of the Dirac equation is the symmetry $ SO(3) \otimes SU(2) $.
The generator of the $ SO(3) $ symmetry group is given by the total momentum
operator, and the generator of $ SU(2) $ group is given by the rotation of the
vector-states in the spinor space, determined by the Dirac, Johnson-Lippmann,
and the new spinor invariants. It is shown that using algebraic approach to the
Dirac problem allows one to calculate the eigenstates and eigenenergies of the
relativistic hydrogen atom and reveals the fundamental role of the principal
quantum number as an independent number, even though it is represented as the
combination of other quantum numbers.
Authors: A. A. Eremko, L. S. Brizhik, V. M. Loktev.
Simulations of biophysical systems have provided a huge contribution to our
fundamental understanding of human physiology and remain a central pillar for
developments in medical devices and human machine interfaces. However, despite
their successes, such simulations usually rely on highly computationally
expensive numerical modelling, which is often inefficient to adapt to new
simulation parameters. This limits their use in dynamic models of human
behavior, for example in modelling the electric fields generated by muscles in
a moving arm. We
argue that transfer learning approaches with conditional generative models are
a viable solution for dynamic simulation with any numerical model.
Simulations of biophysical systems have provided a huge contribution to our
fundamental understanding of human physiology and remain a central pillar for
developments in medical devices and human machine interfaces. However, despite
their successes, such simulations usually rely on highly computationally
expensive numerical modelling, which is often inefficient to adapt to new
simulation parameters. This limits their use in dynamic models of human
behavior, for example in modelling the electric fields generated by muscles in
a moving arm. We propose the alternative approach to use conditional generative
models, which can learn complex relationships between the underlying generative
conditions whilst remaining inexpensive to sample from. As a demonstration of
this concept, we present BioMime, a hybrid architecture that combines elements
of deep latent variable models and conditional adversarial training to
construct a generative model that can both transform existing data samples to
reflect new modelling assumptions and sample new data from a conditioned
distribution. We demonstrate that BioMime can learn to accurately mimic a
complex numerical model of human muscle biophysics and then use this knowledge
to continuously sample from a dynamically changing system in real-time. We
argue that transfer learning approaches with conditional generative models are
a viable solution for dynamic simulation with any numerical model.
Authors: Shihan Ma, Alexander Kenneth Clarke, Kostiantyn Maksymenko, Samuel Deslauriers-Gauthier, Xinjun Sheng, Xiangyang Zhu, Dario Farina.
We construct a 3-variable enrichment of the Lawrence-Krammer-Bigelow (LKB) representation of the braid groups, which is the limit of a pro-nilpotent tower of representations having the original LKB representation as its bottom layer.
We construct a 3-variable enrichment of the Lawrence-Krammer-Bigelow (LKB)
representation of the braid groups, which is the limit of a pro-nilpotent tower
of representations having the original LKB representation as its bottom layer.
We also construct analogous pro-nilpotent towers of representations of surface
braid groups and loop braid groups.
Authors: Martin Palmer, Arthur Soulié.
We show that in
certain geological settings GGN can be significantly mitigated when operating a
multi-gradiometer configuration, which consists of three or more atom
interferometers in the same baseline. Multi-gradiometer experiments, such as
future versions of AION and MAGIS-100, have the potential to probe regions of
scalar ULDM parameter space in the sub-Hz regime that have not been excluded by
existing experiments.
Single-photon atom gradiometry is a powerful experimental technique that can
be employed to search for the oscillation of atomic transition energies induced
by ultralight scalar dark matter (ULDM). In the sub-Hz regime the background is
expected to be dominated by gravity gradient noise (GGN), which arises as a
result of mass fluctuations around the experiment. In this work we model the
GGN as surface Rayleigh waves and construct a likelihood-based analysis that
consistently folds GGN into the sensitivity estimates of vertical atom
gradiometers in the frequency window between 1 mHz and 1 Hz. We show that in
certain geological settings GGN can be significantly mitigated when operating a
multi-gradiometer configuration, which consists of three or more atom
interferometers in the same baseline. Multi-gradiometer experiments, such as
future versions of AION and MAGIS-100, have the potential to probe regions of
scalar ULDM parameter space in the sub-Hz regime that have not been excluded by
existing experiments.
Authors: Leonardo Badurina, Valerie Gibson, Christopher McCabe, Jeremiah Mitchell.
We prove that the resulting system generates a new global process. This statement can be applied to differential equations of various kinds.
Consider the coupling of $2$ evolution equations, each generating a global
process. We prove that the resulting system generates a new global process.
This statement can be applied to differential equations of various kinds. In
particular, it also yields the well posedness of a predator-prey model, where
the coupling is in the differential terms, and of an epidemiological model,
which does not fit previous well posedness results.
Authors: Rinaldo M. Colombo, Mauro Garavello, Matthew Tandy.