Search

Collections of persistently moving active particles are an example of a nonequilibrium heat bath. One way to study the nature of nonequilibrium fluctuations in such systems is to follow the dynamics of an embedded probe particle. With this aim, we study the dynamics of an anisotropic inclusion embedded in a bath of active particles. By studying various statistical correlation functions of the dynamics, we show that the emergent motility of this inclusion depends on its shape as well as the properties of the active bath. We demonstrate that both the decorrelation time of the net force on the inclusion and the dwell time of bath particles in a geometrical trap on the inclusion have a non-monotonic dependence on its shape. We also find that the motility of the inclusion is optimal when the volume fraction of the active bath is close to the value for the onset of motility induced phase separation.

In this paper, we address the detection of co-occurring salient objects (CoSOD) in an image group using frequency statistics in an unsupervised manner, which further enable us to develop a semi-supervised method. While previous works have mostly focused on fully supervised CoSOD, less attention has been allocated to detecting co-salient objects when limited segmentation annotations are available for training. Our simple yet effective unsupervised method US-CoSOD combines the object co-occurrence frequency statistics of unsupervised single-image semantic segmentations with salient foreground detections using self-supervised feature learning. For the first time, we show that a large unlabeled dataset e.g. ImageNet-1k can be effectively leveraged to significantly improve unsupervised CoSOD performance. Our unsupervised model is a great pre-training initialization for our semi-supervised model SS-CoSOD, especially when very limited labeled data is available for training. To avoid propagating erroneous signals from predictions on unlabeled data, we propose a confidence estimation module to guide our semi-supervised training. Extensive experiments on three CoSOD benchmark datasets show that both of our unsupervised and semi-supervised models outperform the corresponding state-of-the-art models by a significant margin (e.g., on the Cosal2015 dataset, our US-CoSOD model has an 8.8% F-measure gain over a SOTA unsupervised co-segmentation model and our SS-CoSOD model has an 11.81% F-measure gain over a SOTA semi-supervised CoSOD model)., Comment: Accepted at IEEE WACV 2024

In recent years, the formation of primordial black holes (PBH) in the early universe inflationary cosmology has garnered significant attention. One plausible scenario for primordial black hole (PBH) formation arises during the preheating stage following inflation. Notably, this scenario does not necessitate any ad-hoc fine-tuning of the scalar field potential. This paper focuses on the growth of primordial density perturbation and the consequent possibility of PBH formation in the preheating stage of the Starobinsky model for inflation. The typical mechanism for PBH formation during preheating is based on the collapse of primordial fluctuations that become super-horizon during inflation (type I) and re-enter the particle horizon in the different phases of cosmic expansion. In this work, we show that there exists a certain range of modes that remain in the sub-horizon (not exited) during inflation (type II modes) but evolve identically to type I modes if they fall into the instability band, leading to large density perturbation above the threshold and can potentially also contribute to the PBH formation. We detail the conditions determining the possible collapse of type I and/or type II modes whose wavelengths are larger than the Jeans length we derive from the effective sound speed of scalar field fluctuations. Since the preheating stage is an 'inflaton' (approximately) matter-dominated phase, we follow the framework of the critical collapse of fluctuations and compute the mass fraction using the well-known Press-Schechter and the Khlopov-Polnarev formalisms, and compare the two. Finally, we comment on the implications of our study for the investigations concerned with primordial accretion and consequent PBH contribution to the dark matter., Comment: 36 pages, 18 figures, version with extended results and discussion

We report on the structural, electrical and magnetic measurements in as-grown polycrystalline samples of Pb10-xCux(PO4)6O. This compound has been recently reported to be a room temperature superconductor. Our as-grown specimen has excellent XRD matching with the original submission of Lee et al. This sample has 1.5% of Cu2S as an impurity phase. A resistive transition around 380 K, possibly corresponding to structural transitions of Cu2S, is observed. No evidence of superconducting to normal state transitions in I-V characteristics at room temperature is obtained. Magnetization measurements show linear diamagnetic behavior that cannot be associated to the superconducting state. Hall measurements provide evidence of hole doping through Cu substitution. In summary, we find no evidence for room temperature ambient pressure superconductivity in Cu doped lead apatite Pb9Cu(PO4)6O.

Domain randomization (DR), which entails training a policy with randomized dynamics, has proven to be a simple yet effective algorithm for reducing the gap between simulation and the real world. However, DR often requires careful tuning of randomization parameters. Methods like Bayesian Domain Randomization (Bayesian DR) and Active Domain Randomization (Adaptive DR) address this issue by automating parameter range selection using real-world experience. While effective, these algorithms often require long computation time, as a new policy is trained from scratch every iteration. In this work, we propose Adaptive Bayesian Domain Randomization via Strategic Fine-tuning (BayRnTune), which inherits the spirit of BayRn but aims to significantly accelerate the learning processes by fine-tuning from previously learned policy. This idea leads to a critical question: which previous policy should we use as a prior during fine-tuning? We investigated four different fine-tuning strategies and compared them against baseline algorithms in five simulated environments, ranging from simple benchmark tasks to more complex legged robot environments. Our analysis demonstrates that our method yields better rewards in the same amount of timesteps compared to vanilla domain randomization or Bayesian DR.

Holes in silicon quantum dots are promising for spin qubit applications due to the strong intrinsic spin-orbit coupling. The spin-orbit coupling produces complex hole-spin dynamics, providing opportunities to further optimize spin qubits. Here, we demonstrate a singlet-triplet qubit using hole states in a planar metal-oxide-semiconductor double quantum dot. We observe rapid qubit control with singlet-triplet oscillations up to 400 MHz. The qubit exhibits promising coherence, with a maximum dephasing time of 600 ns, which is enhanced to 1.3 us using refocusing techniques. We investigate the magnetic field anisotropy of the eigenstates, and determine a magnetic field orientation to improve the qubit initialisation fidelity. These results present a step forward for spin qubit technology, by implementing a high quality singlet-triplet hole-spin qubit in planar architecture suitable for scaling up to 2D arrays of coupled qubits.