The Variable Energy Cyclotron Centre (VECC) is a research and development unit of the Department of Atomic Energy. The VECC is located in Kolkata, India and performs research in basic and applied nuclear sciences and development of the latest nuclear particle accelerators. It has a collaboration with the European Organization for Nuclear Research.The Centre houses a 224 cm cyclotron—the first of its kind in India—which has been operational since 16 June 1977. It provides proton, deuteron, alpha particle and heavy ion beams of various energies to other institutions.The Centre consists of major facilities such as K130 Cyclotron, K500 Superconducting Cyclotron, Cyclone-30 Medical Cyclotron, Radioactive Ion Beam (RIB) Facility, Computing Centre, Regional Radiation Medicine Centre and a new Campus for the proposed ANURIB project at New Town, Rajarhat. The ANURIB (Advanced National facility for Unstable & Rare-Isotope Beams) is a planned facility, to be constructed in collaboration with the Canada-based research institute TRIUMF. ANURIB is going to conduct experiments of unstable & rare isotope beams.
A remarkably high value of specific capacitance of 450 F/g has been observed through electrochemical measurements in the electrode made of multiferroic Bismuth Ferrite (BFO) in the form of nanorods protruding out. These BFO nanorods were developed on porous Anodised Alumina (AAO) templates using wet chemical technique. Diameters of nanorods were in the range of 20-100 nm. The high capacitance is attributed to the nanostructure. The active surface charge has been evaluated electrochemically by cyclic voltammetry (CV) at different scanning rates and charge-discharge studies. The specific capacitances were constant after several cycles of charge-discharge leading to their useful application in devices. The mechanism of accumulation of charge on the electrode surface has been studied.
The Compressed Baryonic Matter (CBM) experiment at the future Facility for Antiproton and Ion Research (FAIR) is a heavy-ion experiment designed to study nuclear matter at the highest baryonic density. For high-statistics measurements of rare probes, collision rates of up to 10 MHz are targeted. The experiment, therefore, requires fast and radiation-hard detectors, self-triggered detector front-ends, free-streaming readout architecture, and online event reconstruction. The Silicon Tracking System (STS) is the main tracking detector of CBM, designed to reconstruct the trajectories of charged particles with efficiency larger than 95%, a relative momentum uncertainty better than 2% for particle momenta larger than 1 GeV/c inside a 1Tm magnetic field, and to identify complex decay topologies. It comprises 876 double-sided silicon strip modules arranged in 8 tracking stations. A prototype of this detector, consisting of 12 modules arranged in three tracking stations, is installed in the mini-CBM demonstrator. This experimental setup is a small-scale precursor to the full CBM detector, composed of sub-units of all major CBM systems installed on the SIS18 beamline. In various beam campaigns taken between 2021 and 2024, heavy ion collisions at 1-2AGeV with an average collision rate of 500kHz have been recorded. This allows for the evaluation of the operational performance of the STS detector, including signal-to-noise ratio, charge distribution, time and position resolution, hit reconstruction efficiency, and its potential for track and vertex reconstruction.
Lightweight convolutional and transformer-based networks are increasingly used for real-time image classification on resource-constrained hardware, yet their practical performance is highly sensitive to training hyperparameters. This work systematically quantifies how controlled hyperparameter choices affect both accuracy and deployability for seven modern lightweight backbones-ConvNeXt-Tiny, EfficientNetV2-S, MobileNetV3-L, MobileViT v2 (S/XS), RepVGG-A2, and TinyViT-21M trained from scratch on a class-balanced 90K/10K subset of ImageNet-1K under a standardized 300-epoch protocol. We isolate the effects of learning-rate magnitude and cosine scheduling, optimizer selection (SGD vs. AdamW where appropriate), and progressively stronger regularization via RandAugment, Mixup, CutMix, and label smoothing, complemented by constrained automated searches (Optuna and population-based training). Beyond training-time analysis, we add a deployment-focused evaluation: inference latency and throughput are benchmarked on an NVIDIA L40s GPU across batch sizes 1-512, and edge feasibility is examined via Edge CPU Platform under sustained workloads. Results show that hyperparameter tuning without architectural modification yields consistent accuracy gains ([Formula: see text] Top-1 over baseline) and reveals architecture-dependent stability regions. Several models deliver strong real-time operating points: MobileNetV3-L and RepVGG-A2 achieve very low latency with high throughput on GPU, while edge tests highlight the limited benefit of batching on low-power CPUs and the importance of latency-centric model choice. The code and logs may be seen at: https://github.com/VineetKumarRakesh/lcnn-opt .
Diffusion models have recently advanced photorealistic human synthesis, although practical talking-head generation (THG) remains constrained by high inference latency, temporal instability such as flicker and identity drift, and imperfect audio-visual alignment under challenging speech conditions. This paper introduces TempoSyncDiff, a reference-conditioned latent diffusion framework that explores few-step inference for efficient audio-driven talking-head generation. The approach adopts a teacher-student distillation formulation in which a diffusion teacher trained with a standard noise prediction objective guides a lightweight student denoiser capable of operating with significantly fewer inference steps to improve generation stability. The framework incorporates identity anchoring and temporal regularization designed to mitigate identity drift and frame-to-frame flicker during synthesis, while viseme-based audio conditioning provides coarse lip motion control. Experiments on the LRS3 dataset report denoising-stage component-level metrics relative to VAE reconstructions and preliminary latency characterization, including CPU-only and edge computing measurements and feasibility estimates for edge deployment. The results suggest that distilled diffusion models can retain much of the reconstruction behaviour of a stronger teacher while enabling substantially lower latency inference. The study is positioned as an initial step toward practical diffusion-based talking-head generation under constrained computational settings. GitHub: https://mazumdarsoumya.github.io/TempoSyncDiff
In this work, we have calculated the transport coefficients: shear viscosity and thermal conductivity inside the neutron star core. Our calculation is based on the relativistic kinetic theory approach using a modified BUU equation for quasi-particles whose mass and the chemical-potential and thus in turn the Fermi surface varies with the baryonic density rho B and the temperature of the medium. We have used the relaxation time approximation. For the description of the hadronic matter inside the neutron star, we consider the relativistic mean field model with three different kinds of parameterizations. We have found that the shear viscosity is predominantly influenced by neutrons, while thermal conductivity is primarily dominated by electrons.
