This study explores the impact of Islamophobia on Muslim children in Australia, shedding light on their lived experiences and the long-term consequences of anti-Muslim hostility. Drawing on 28 semi-structured interviews and four focus group discussions (FGD) conducted with Muslim parents, predominantly mothers, across New South Wales and Victoria, the study highlights the ways in which children are perceived as threats, often facing harassment in both formal settings, such as schools, and informal public spaces. The research exposes how institutional Islamophobia from government and media sources, creates a climate for individual experiences of Islamophobia identified by the parents. Overall, the findings reveal that Islamophobic discourse places a significant psychological burden on Muslim children, impacting their sense of belonging, self-esteem, and emotional well-being. The rhetoric of the “dangerous Muslim” extends to children, with instances of bullying, name-calling, and harassment contributing to children’s feelings of alienation and insecurity. Parents often experience heightened anxiety, which children internalize, further embedding the cycle of fear and exclusion that propels parents to adopt strategies to protect their children from emotional harm. The paper argues that Islamophobia is not an isolated issue but a systemic societal challenge that requires coordinated efforts to dismantle structural racism and Islamophobia. It calls for long-term, multi-stakeholder approaches.
Online 3D reconstruction requires estimating camera pose and scene geometry under strict causal and bounded-memory constraints. Existing methods often suffer from drift, jitter, or collapse on long sequences. We trace these failures to a fundamental mismatch. Streaming geometry is inherently temporally heterogeneous, with evidence ranging from short-lived correspondences to persistent global scale. However, current architectures impose uniform and pathological influence patterns. For example, sliding windows enforce hard cutoffs, while ungated recurrence and causal attention cause cache saturation and spike-like attention sinks. To resolve this, we formalize geometric propagation as an evidence influence kernel and propose HorizonStream, a long-horizon Transformer that explicitly factorizes this kernel. For the long-range temporal factor, Geometric Linear Attention learns channel-wise decay rates to enable bounded, multi-timescale propagation of geometric evidence. For the short-range spatial factor, Geometric Local Attention with Spatiotemporal RoPE performs reliable 3D matching while suppressing attention sinks. Finally, Metric Readout Tokens recover stable scale and rigid pose directly from the persistent geometric state. Extensive experiments show that HorizonStream, trained on only 48-frame clips, generalizes stably to sequences exceeding 10,000 frames with constant memory and linear time, achieving state-of-the-art streaming 3D reconstruction performance. Project Page: https://3dagentworld.github.io/horizonstream/
Understanding long videos with multimodal large language models (MLLMs) requires selecting a compact set of frames from thousands of candidates, yet identifying the right frames seemingly requires understanding the video first. We resolve this circular dependency with a simple observation: cross-modal attention at validation-selected extraction layers in MLLMs already provides query-relevant frame evidence without requiring autoregressive generation. We exploit this property to build DAFS (Dynamic Attention-based Budget-aware Frame Selection), a training-free frame selector. A lightweight MLLM selector, even with only 2B parameters, can extract frame-level evidence by converting selected-layer attention into relevance scores through query-conditioned aggregation. This enables cross-frame comparison without autoregressive decoding. To handle the selector's own context constraint, we formulate the joint allocation of candidate pool size and per-frame token budget as a discrete optimization problem solved by dynamic programming. Under a 32-frame budget, our selector improves over uniform sampling by up to 6.4 points on Video-MME and outperforms prior training-based selectors under matched frame budgets, while generalizing across selector and answerer backbones, and across tasks, without retraining.
Abstract The Deep Underground Neutrino Experiment (DUNE) Far Detector (FD) Photon Detection System (PDS) employs the X-Arapuca concept, a photon trapping system relying on reflective surfaces and dichroic filters. In this paper are reported measurements, performed at the University of Milano-Bicocca, aimed at increasing the FD Horizontal Drift (HD) PDS module efficiency. The baseline implementation of the X-Arapuca concept for the FD-HD PDS module is close to the DUNE requirements as demonstrated in the collaborations laboratory testing. However, an increased performance would provide a safety margin for a detector planned to be operated for 30 years, without possibility of performing maintenance. A higher detector performance would also benefit the DUNE low energy physics program. The already proven Milano-Bicocca setup has been utilized to test different PDS module configurations comparing them to the original baseline. Exploiting prior knowledge of the X-Arapuca components and Geant4 based optical simulations it has been possible to achieve up to an $$\sim $$ ∼ 84% performance increase over the baseline design. In the following it is presented the testing procedure, the performed measurements and a brief discussion on the obtained results.
Diffusion models have achieved remarkable success in image generation, yet their deployment remains constrained by the heavy computational cost and the need for numerous inference steps. Previous efforts on fewer-step distillation attempt to skip redundant steps by training compact student models, yet they often suffer from heavy retraining costs and degraded generalization. In this work, we take a different perspective: we accelerate smartly, not evenly, applying smaller speedups to early semantic stages and larger ones to later redundant phases. We instantiate this phase-aware strategy with two experts that specialize in slow and fast denoising phases. Surprisingly, instead of investing massive effort in retraining student models, we find that simply equipping the base model with lightweight LoRA adapters achieves both efficient acceleration and strong generalization. We refer to these two adapters as Slow-LoRA and Fast-LoRA. Through extensive experiments, our method achieves up to 5 × acceleration over the base model while maintaining comparable visual quality across diverse benchmarks. Remarkably, the LoRA experts are trained with only 1 samples on a single V100 within one hour, yet the resulting models generalize strongly on unseen prompts. Code is available at https://zhuobaidong.github.io/Glance/ .
