Tootfinder

Hah, RNZ Jesse Mulligan is interviewing another scientist I know, former colleague Steve Pawson - the second in a week (last week it was Nicola Sullivan re spiders)... That's Aotearoa for ya.

This is how we celebrate the 4th of July here in Aotearoa... With an Ōtautahi-brewed America Pale Ale, and gratitude that we took the precaution of drifting to the other end of the earth over the last few hundred million years.

Looking at RNZ's schedule for today, I'm flabbergasted (as I am most days) with how obsessed the media are with 'AI' this and that and everything. To the exclusion of everything else. The hype is at fever pitch, and the masses are getting no wiser as a result. It's idiocy.

MobileIE: An Extremely Lightweight and Effective ConvNet for Real-Time Image Enhancement on Mobile Devices
Hailong Yan, Ao Li, Xiangtao Zhang, Zhe Liu, Zenglin Shi, Ce Zhu, Le Zhang
https://arxiv.org/abs/2507.01838

MobileIE: An Extremely Lightweight and Effective ConvNet for Real-Time Image Enhancement on Mobile Devices
Recent advancements in deep neural networks have driven significant progress in image enhancement (IE). However, deploying deep learning models on resource-constrained platforms, such as mobile devices, remains challenging due to high computation and memory demands. To address these challenges and facilitate real-time IE on mobile, we introduce an extremely lightweight Convolutional Neural Network (CNN) framework with around 4K parameters. Our approach integrates reparameterization with an Incr…

Oh my word, what the hell are we doing? #libre

Just found out that an old intranet we initially created for a customer back in 2008-9 (on Drupal 4.7 or maybe 5, subsequently upgraded through to Drupal 6) is still running (not on the open internet). Their MS Sharepoint replacement for it has been... less than stunningly successful in its designated role.
I suspect that original Intranet, even inflation adjusted, would've cost the orders of magnitude less in up-front development and per-year maintenance.

All You Need Is Binary Search! A Practical View on Lightweight Database Indexing on GPUs
Justus Henneberg, Felix Schuhknecht
https://arxiv.org/abs/2506.01576

All You Need Is Binary Search! A Practical View on Lightweight Database Indexing on GPUs
Performing binary search on a sorted dense array is a widely used baseline when benchmarking sophisticated index structures: It is simple, fast to build, and indexes the dataset with minimal memory footprint. However, the popular opinion is that it cannot compete with sophisticated indexes in terms of lookup performance, and hence, should not actually be considered in practice. Interestingly, in our recent works on (even more sophisticated) GPU-resident index structures, we observed the surpr…

"EV battery recycling key to easing future lithium supply constraints, modeling study shows"
#EV #ElectricVehicle #Cars

EV battery recycling key to easing future lithium supply constraints, modeling study shows
Lightweight, powerful lithium-ion batteries are crucial for the transition to electric vehicles, and global demand for lithium is set to grow rapidly over the next 25 years. A new analysis from the University of California, Davis, published May 29 in Nature Sustainability, looks at how new mining operations and battery recycling could meet that demand. Recycling could play a big role in easing supply constraints, the researchers found.

WishGI: Lightweight Static Global Illumination Baking via Spherical Harmonics Fitting
Junke Zhu, Zehan Wu, Qixing Zhang, Cheng Liao, Zhangjin Huang
https://arxiv.org/abs/2506.01288

WishGI: Lightweight Static Global Illumination Baking via Spherical Harmonics Fitting
Global illumination combines direct and indirect lighting to create realistic lighting effects, bringing virtual scenes closer to reality. Static global illumination is a crucial component of virtual scene rendering, leveraging precomputation and baking techniques to significantly reduce runtime computational costs. Unfortunately, many existing works prioritize visual quality by relying on extensive texture storage and massive pixel-level texture sampling, leading to large performance overhead.…

Research on Improving the High Precision and Lightweight Diabetic Retinopathy Detection of YOLOv8n
Fei Yuhuan, Sun Xufei, Zang Ran, Wang Gengchen, Su Meng, Liu Fenghao
https://arxiv.org/abs/2507.00780

Research on Improving the High Precision and Lightweight Diabetic Retinopathy Detection of YOLOv8n
Early detection and diagnosis of diabetic retinopathy is one of the current research focuses in ophthalmology. However, due to the subtle features of micro-lesions and their susceptibility to background interference, ex-isting detection methods still face many challenges in terms of accuracy and robustness. To address these issues, a lightweight and high-precision detection model based on the improved YOLOv8n, named YOLO-KFG, is proposed. Firstly, a new dynamic convolution KWConv and C2f-KW mod…

HCVR: A Hybrid Approach with Correlation-aware Voting Rules for Feature Selection
Nikita Bhedasgaonkar, Rushikesh K. Joshi
https://arxiv.org/abs/2507.02073

HCVR: A Hybrid Approach with Correlation-aware Voting Rules for Feature Selection
In this paper, we propose HCVR (Hybrid approach with Correlation-aware Voting Rules), a lightweight rule-based feature selection method that combines Parameter-to-Parameter (P2P) and Parameter-to-Target (P2T) correlations to eliminate redundant features and retain relevant ones. This method is a hybrid of non-iterative and iterative filtering approaches for dimensionality reduction. It is a greedy method, which works by backward elimination, eliminating possibly multiple features at every step.…

Geminet: Learning the Duality-based Iterative Process for Lightweight Traffic Engineering in Changing Topologies
Ximeng Liu, Shizhen Zhao, Xinbing Wang
https://arxiv.org/abs/2506.23640

Geminet: Learning the Duality-based Iterative Process for Lightweight Traffic Engineering in Changing Topologies
Recently, researchers have explored ML-based Traffic Engineering (TE), leveraging neural networks to solve TE problems traditionally addressed by optimization. However, existing ML-based TE schemes remain impractical: they either fail to handle topology changes or suffer from poor scalability due to excessive computational and memory overhead. To overcome these limitations, we propose Geminet, a lightweight and scalable ML-based TE framework that can handle changing topologies. Geminet is built…

Exploiting complex 3D-printed surface structures for portable quantum technologies
Nathan Cooper, David Johnson, Benjamin Hopton, Matthew Overton, David Stupple, Alexandra Bratu, Edward Wilson, John Robinson, Laurence Coles, Manolis Papastavrou, Lucia Hackermueller
https://arxiv.org/abs/2507.01476

Exploiting complex 3D-printed surface structures for portable quantum technologies
Portable quantum technologies require robust, lightweight apparatus with superior performance. For techniques dependent upon high-vacuum environments, such as atom interferometers and atomic clocks, 3D-printing enables new avenues to tailor in-vacuum gas propagation dynamics. We demonstrate intricate, fine-scale surface patterning of 3D-printed vacuum components to increase the rate at which gas particles collide with the surface. By applying a non-evaporable getter coating for use as a surface…

Flying Co-Stereo: Enabling Long-Range Aerial Dense Mapping via Collaborative Stereo Vision of Dynamic-Baseline
Zhaoying Wang, Xingxing Zuo, Wei Dong
https://arxiv.org/abs/2506.00546

Flying Co-Stereo: Enabling Long-Range Aerial Dense Mapping via Collaborative Stereo Vision of Dynamic-Baseline
Lightweight long-range mapping is critical for safe navigation of UAV swarms in large-scale unknown environments. Traditional stereo vision systems with fixed short baselines face limited perception ranges. To address this, we propose Flying Co-Stereo, a cross-agent collaborative stereo vision system that leverages the wide-baseline spatial configuration of two UAVs for long-range dense mapping. Key innovations include: (1) a dual-spectrum visual-inertial-ranging estimator for robust baseline e…

Exploring metrics for analyzing dynamic behavior in MPI programs via a coupled-oscillator model
Ayesha Afzal, Georg Hager, Gerhard Wellen
https://arxiv.org/abs/2506.02792

Exploring metrics for analyzing dynamic behavior in MPI programs via a coupled-oscillator model
We propose a novel, lightweight, and physically inspired approach to modeling the dynamics of parallel distributed-memory programs. Inspired by the Kuramoto model, we represent MPI processes as coupled oscillators with topology-aware interactions, custom coupling potentials, and stochastic noise. The resulting system of nonlinear ordinary differential equations opens a path to modeling key performance phenomena of parallel programs, including synchronization, delay propagation and decay, bottle…

This https://arxiv.org/abs/2505.23655 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csCR_…

Keyed Chaotic Tensor Transformations for Secure And Attributable Neural Inference
This work introduces a novel framework for secure and privacy-preserving neural network inference based on keyed chaotic dynamical transformations. The proposed method applies a deterministic, cryptographically seeded chaotic system to tensors, producing non-invertible, user-specific transformations that enable authenticated inference, tensor-level watermarking, and data attribution. This framework offers a scalable and lightweight alternative to conventional cryptographic techniques, and estab…

This https://arxiv.org/abs/2504.10545 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csIR_…

HSTU-BLaIR: Lightweight Contrastive Text Embedding for Generative Recommender
Recent advances in recommender systems have underscored the complementary strengths of generative modeling and pretrained language models. We propose HSTU-BLaIR, a hybrid framework that augments the Hierarchical Sequential Transduction Unit (HSTU)-based generative recommender with BLaIR, a lightweight contrastive text embedding model. This integration enriches item representations with semantic signals from textual metadata while preserving HSTU's powerful sequence modeling capabilities. We e…

Online Audio-Visual Autoregressive Speaker Extraction
Zexu Pan, Wupeng Wang, Shengkui Zhao, Chong Zhang, Kun Zhou, Yukun Ma, Bin Ma
https://arxiv.org/abs/2506.01270

Online Audio-Visual Autoregressive Speaker Extraction
This paper proposes a novel online audio-visual speaker extraction model. In the streaming regime, most studies optimize the audio network only, leaving the visual frontend less explored. We first propose a lightweight visual frontend based on depth-wise separable convolution. Then, we propose a lightweight autoregressive acoustic encoder to serve as the second cue, to actively explore the information in the separated speech signal from past steps. Scenario-wise, for the first time, we study ho…

A little tip to all those folks using Amazon Web Services or Microsoft Azure or Google Cloud: you're paying ~ 20x more than you need to be paying. If you claim you're using 'special services only provided by __insert name of ripoff cloud provider here__' then you've got your organisation locked into a single provider with the inability to move to a better one. Consider your options very carefully & build vendor-agnostic systems. Dependence on

Phase-based analysis and control of low Reynolds number aeroelastic flows
Chathura R. Sumanasiri, Tulsi Ram Sahu, Aditya G. Nair
https://arxiv.org/abs/2506.00306

Phase-based analysis and control of low Reynolds number aeroelastic flows
Flutter in lightweight airfoils under unsteady flows presents a critical challenge in aeroelastic stability and control. This study uncovers phase-localized mechanisms that drive the onset and suppression of flutter in a freely pitching airfoil at low Reynolds number. By introducing targeted impulsive stiffness perturbations, we identify critical phases that trigger instability. Using phase-sensitivity functions, energy-transfer metrics, and dynamic mode decomposition, we show that flutter aris…

Currently working on implementing Matrix's new 'Matrix Authentication Service' for our NZ-based instance... non-trivial. But it's such an impressive system (it's doing far more than competitors like Salesforce Slack and Microsoft Teams).

Meta-heuristic design of a light-weight homologous backup structure of the primary reflector for the Large Submillimeter Telescope
Chihiro Imamura, Yoichi Tamura, Hiroaki Kawamura, Toshiaki Kimura, Akio Taniguchi, Mikio Kurita
#toXiv_bot_toot

He makes a very solid case: https://yewtu.be/watch?v=bNTg9EX7MLw or https://www.youtube.com/watch?v=bNTg9EX7ML

Explainable AI for Comprehensive Risk Assessment for Financial Reports: A Lightweight Hierarchical Transformer Network Approach
Xue Wen Tan, Stanley Kok
https://arxiv.org/abs/2506.23767

Explainable AI for Comprehensive Risk Assessment for Financial Reports: A Lightweight Hierarchical Transformer Network Approach
Every publicly traded U.S. company files an annual 10-K report containing critical insights into financial health and risk. We propose Tiny eXplainable Risk Assessor (TinyXRA), a lightweight and explainable transformer-based model that automatically assesses company risk from these reports. Unlike prior work that relies solely on the standard deviation of excess returns (adjusted for the Fama-French model), which indiscriminately penalizes both upside and downside risk, TinyXRA incorporates ske…

This https://arxiv.org/abs/2505.14667 has been replaced.
initial toot: https://mastoxiv.page/@arXiv_csAI_…

SAFEPATH: Preventing Harmful Reasoning in Chain-of-Thought via Early Alignment
Large Reasoning Models (LRMs) have become powerful tools for complex problem solving, but their structured reasoning pathways can lead to unsafe outputs when exposed to harmful prompts. Existing safety alignment methods reduce harmful outputs but can degrade reasoning depth, leading to significant trade-offs in complex, multi-step tasks, and remain vulnerable to sophisticated jailbreak attacks. To address this, we introduce SAFEPATH, a lightweight alignment method that fine-tunes LRMs to emit a…

Snap aims to ship lightweight, consumer AR glasses called Specs in 2026, lighter than its Spectacles 5 glasses for devs, but with many of the same capabilities (Maxwell Zeff/TechCrunch)
https://techcrunch.com/2025/06/10/snap-plans-to-s…

Snap plans to sell lightweight, consumer AR glasses in 2026 | TechCrunch
Snap is back with a pair of lightweight, AR smart glasses. For the first time in years, the company wants to sell them to consumers.

Humanoid World Models: Open World Foundation Models for Humanoid Robotics
Muhammad Qasim Ali, Aditya Sridhar, Shahbuland Matiana, Alex Wong, Mohammad Al-Sharman
https://arxiv.org/abs/2506.01182

Humanoid World Models: Open World Foundation Models for Humanoid Robotics
Humanoid robots have the potential to perform complex tasks in human centered environments but require robust predictive models to reason about the outcomes of their actions. We introduce Humanoid World Models (HWM) a family of lightweight open source video based models that forecast future egocentric observations conditioned on actions. We train two types of generative models Masked Transformers and FlowMatching on 100 hours of humanoid demonstrations. Additionally we explore architectural var…

Discretely structured magnetic flux concentrators
Yujun Shi, Xiaoting Feng
https://arxiv.org/abs/2506.01424 https://arxiv.org/pdf/250…

Discretely structured magnetic flux concentrators
Conventional magnetic flux concentrators (MFCs) are typically designed with solid structures, which may not be optimal for applications requiring lightweight design or material efficiency. In this study, we investigate the feasibility of spatially discretized MFCs in guiding and concentrating magnetic flux, through finite element simulations. Our results demonstrate that discretely structured MFCs can achieve performance comparable to that of solid counterparts while significantly reducing mate…

Improved belief propagation is sufficient for real-time decoding of quantum memory
Tristan M\"uller, Thomas Alexander, Michael E. Beverland, Markus B\"uhler, Blake R. Johnson, Thilo Maurer, Drew Vandeth
https://arxiv.org/abs/2506.01779

Improved belief propagation is sufficient for real-time decoding of quantum memory
We introduce a new heuristic decoder, Relay-BP, targeting real-time quantum circuit decoding for large-scale quantum computers. Relay-BP achieves high accuracy across circuit-noise decoding problems: significantly outperforming BP+OSD+CS-10 for bivariate-bicycle codes and comparable to min-weight-matching for surface codes. As a lightweight message-passing decoder, Relay-BP is inherently parallel, enabling rapid low-footprint decoding with FPGA or ASIC real-time implementations, similar to stan…

Lightweight Fingernail Haptic Device: Unobstructed Fingerpad Force and Vibration Feedback for Enhanced Virtual Dexterous Manipulation
Yunxiu Xu, Siyu Wang, Shoichi Hasegawa
https://arxiv.org/abs/2506.21417

Lightweight Fingernail Haptic Device: Unobstructed Fingerpad Force and Vibration Feedback for Enhanced Virtual Dexterous Manipulation
This study presents a lightweight, wearable fingertip haptic device that provides physics-based haptic feedback for dexterous manipulation in virtual environments without hindering real-world interactions. The device, designed with thin strings and actuators attached to the fingernails, ensures minimal weight (1.55 g per finger) and preserves finger flexibility. Integrating the software with a physics engine renders multiple types of haptic feedback (grip force, collision, and sliding vibration…

Enable Lightweight and Precision-Scalable Posit/IEEE-754 Arithmetic in RISC-V Cores for Transprecision Computing
Qiong Li, Chao Fang, Longwei Huang, Jun Lin, Zhongfeng Wang
https://arxiv.org/abs/2505.19096

Enable Lightweight and Precision-Scalable Posit/IEEE-754 Arithmetic in RISC-V Cores for Transprecision Computing
While posit format offers superior dynamic range and accuracy for transprecision computing, its adoption in RISC-V processors is hindered by the lack of a unified solution for lightweight, precision-scalable, and IEEE-754 arithmetic compatible hardware implementation. To address these challenges, we enhance RISC-V processors by 1) integrating dedicated posit codecs into the original FPU for lightweight implementation, 2) incorporating multi/mixed-precision support with dynamic exponent size for…

Multi-Branch DNN and CRLB-Ratio-Weight Fusion for Enhanced DOA Sensing via a Massive H$^2$AD MIMO Receiver
Feng Shu, Jiatong Bai, Di Wu, Wei Zhu, Bin Deng, Fuhui Zhou, Jiangzhou Wang
https://arxiv.org/abs/2506.23203

Multi-Branch DNN and CRLB-Ratio-Weight Fusion for Enhanced DOA Sensing via a Massive H$^2$AD MIMO Receiver
As a green MIMO structure, massive H$^2$AD is viewed as a potential technology for the future 6G wireless network. For such a structure, it is a challenging task to design a low-complexity and high-performance fusion of target direction values sensed by different sub-array groups with fewer use of prior knowledge. To address this issue, a lightweight Cramer-Rao lower bound (CRLB)-ratio-weight fusion (WF) method is proposed, which approximates inverse CRLB of each subarray using antenna number r…

Evidence for a sub-Jovian planet in the young TWA 7 #disk : #Planet Imaged by NASA #Webb Is Lightest Ever Seen: https://webbtelescope.org/contents/news-releases/2025/news-2025-126 and https://www.esa.int/Science_Exploration/Space_Science/Webb/Webb_finds_evidence_of_a_lightweight_planet_around_TWA_7

This https://arxiv.org/abs/2409.10289 has been replaced.
link: https://scholar.google.com/scholar?q=a

ReflectDiffu:Reflect between Emotion-intent Contagion and Mimicry for Empathetic Response Generation via a RL-Diffusion Framework
Empathetic response generation necessitates the integration of emotional and intentional dynamics to foster meaningful interactions. Existing research either neglects the intricate interplay between emotion and intent, leading to suboptimal controllability of empathy, or resorts to large language models (LLMs), which incur significant computational overhead. In this paper, we introduce ReflectDiffu, a lightweight and comprehensive framework for empathetic response generation. This framework inc…

S\"{o}ze: One Network Telemetry Is All You Need for Per-flow Weighted Bandwidth Allocation at Scale
Weitao Wang, T. S. Eugene Ng
https://arxiv.org/abs/2506.00834

Söze: One Network Telemetry Is All You Need for Per-flow Weighted Bandwidth Allocation at Scale
Weighted bandwidth allocation is a powerful abstraction that has a wide range of use cases in modern data center networks. However, realizing highly agile and precise weighted bandwidth allocation for large-scale cloud environments is fundamentally challenging. In this paper, we propose Söze, a lightweight decentralized weighted bandwidth allocation system that leverages simple network telemetry features of commodity Ethernet switches. Given the flow weights, Söze can effectively use the tele…

Ok, this is hilarious. https://www.yankodesign.com/2025/06/27/worlds-narrowest-fiat-panda-is-one-anorexic-19-inch-wide-ev-destined-for-the-record-books/

RaGNNarok: A Light-Weight Graph Neural Network for Enhancing Radar Point Clouds on Unmanned Ground Vehicles
David Hunt, Shaocheng Luo, Spencer Hallyburton, Shafii Nillongo, Yi Li, Tingjun Chen, Miroslav Pajic
https://arxiv.org/abs/2507.00937

RaGNNarok: A Light-Weight Graph Neural Network for Enhancing Radar Point Clouds on Unmanned Ground Vehicles
Low-cost indoor mobile robots have gained popularity with the increasing adoption of automation in homes and commercial spaces. However, existing lidar and camera-based solutions have limitations such as poor performance in visually obscured environments, high computational overhead for data processing, and high costs for lidars. In contrast, mmWave radar sensors offer a cost-effective and lightweight alternative, providing accurate ranging regardless of visibility. However, existing radar-base…

We've been studying the cognitive science of *diagrams*, and have recently put out a new (very) lightweight programming language and tool for diagramming models in formal methods. Check it out here!
https://blog.brownplt.org/2025/06/09/copeanddrag.html

MuseControlLite: Multifunctional Music Generation with Lightweight Conditioners
Fang-Duo Tsai, Shih-Lun Wu, Weijaw Lee, Sheng-Ping Yang, Bo-Rui Chen, Hao-Chung Cheng, Yi-Hsuan Yang
https://arxiv.org/abs/2506.18729

MuseControlLite: Multifunctional Music Generation with Lightweight Conditioners
We propose MuseControlLite, a lightweight mechanism designed to fine-tune text-to-music generation models for precise conditioning using various time-varying musical attributes and reference audio signals. The key finding is that positional embeddings, which have been seldom used by text-to-music generation models in the conditioner for text conditions, are critical when the condition of interest is a function of time. Using melody control as an example, our experiments show that simply adding …

Logit-Gap Steering: Efficient Short-Suffix Jailbreaks for Aligned Large Language Models
Tung-Ling Li, Hongliang Liu
https://arxiv.org/abs/2506.24056 https:…

Logit-Gap Steering: Efficient Short-Suffix Jailbreaks for Aligned Large Language Models
We introduce logit-gap steering, a fast jailbreak framework that casts the refusal-affirmation gap of RLHF-aligned language models as a single pass over the vocabulary. A forward-computable score blends gap reduction with lightweight proxies for KL penalty and reward shift, allowing a "sort-sum-stop" sweep to complete in under a second and return a short suffix--two orders of magnitude fewer model calls than beam or gradient attacks. The same suffix generalises to unseen prompts and scales from…

Extreme Learning Machines for Exoplanet Simulations: A Faster, Lightweight Alternative to Deep Learning
Tara P. A. Tahseen, Lu\'is F. Sim\~oes, Kai Hou Yip, Nikolaos Nikolaou, Jo\~ao M. Mendon\c{c}a, Ingo P. Waldmann
https://arxiv.org/abs/2506.19679

Extreme Learning Machines for Exoplanet Simulations: A Faster, Lightweight Alternative to Deep Learning
Increasing resolution and coverage of astrophysical and climate data necessitates increasingly sophisticated models, often pushing the limits of computational feasibility. While emulation methods can reduce calculation costs, the neural architectures typically used--optimised via gradient descent--are themselves computationally expensive to train, particularly in terms of data generation requirements. This paper investigates the utility of the Extreme Learning Machine (ELM) as a lightweight, no…

Working again on a project with Delphi 1.0 was a pleasure. Few IDEs nowadays are as simple, lightweight, complete, useful, and stable as Delphi was in… 1995.
Le sigh.
https://akos.ma/blog/conway-in-borland-delphi/

Conway in Borland Delphi
Yet another version of my Polyglot Conway project, this time in good old Borland Delphi 1.0 (which recently celebrated its 30th anniversary!) from 1995, running on Windows 3.1, and as usual you can found the source code on my GitLab account.

A Reinforcement Learning-Based Telematic Routing Protocol for the Internet of Underwater Things
Mohammadhossein Homaei, Mehran Tarif, Agustin Di Bartolo, Oscar Mogollon Gutierrez, Mar Avila
https://arxiv.org/abs/2506.00133

A Reinforcement Learning-Based Telematic Routing Protocol for the Internet of Underwater Things
The Internet of Underwater Things (IoUT) faces major challenges such as low bandwidth, high latency, mobility, and limited energy resources. Traditional routing protocols like RPL, which were designed for land-based networks, do not perform well in these underwater conditions. This paper introduces RL-RPL-UA, a new routing protocol that uses reinforcement learning to improve performance in underwater environments. Each node includes a lightweight RL agent that selects the best parent node based…

Lightweight Multi-Frame Integration for Robust YOLO Object Detection in Videos
Yitong Quan, Benjamin Kiefer, Martin Messmer, Andreas Zell
https://arxiv.org/abs/2506.20550

Lightweight Multi-Frame Integration for Robust YOLO Object Detection in Videos
Modern image-based object detection models, such as YOLOv7, primarily process individual frames independently, thus ignoring valuable temporal context naturally present in videos. Meanwhile, existing video-based detection methods often introduce complex temporal modules, significantly increasing model size and computational complexity. In practical applications such as surveillance and autonomous driving, transient challenges including motion blur, occlusions, and abrupt appearance changes can …

Ok #Fediverse: if I was keen to bootstrap an Aotearoa-based tech services #cooperative, what coop models would people suggest I look at - looking for good descriptions of how they'd be set up, especially to recognise the contributions of those involved. All insights appreciated! I'…

Lightweight Physics-Informed Zero-Shot Ultrasound Plane Wave Denoising
Hojat Asgariandehkordi, Mostafa Sharifzadeh, Hassan Rivaz
https://arxiv.org/abs/2506.21499

Lightweight Physics-Informed Zero-Shot Ultrasound Plane Wave Denoising
Ultrasound Coherent Plane Wave Compounding (CPWC) enhances image contrast by combining echoes from multiple steered transmissions. While increasing the number of angles generally improves image quality, it drastically reduces the frame rate and can introduce blurring artifacts in fast-moving targets. Moreover, compounded images remain susceptible to noise, particularly when acquired with a limited number of transmissions. We propose a zero-shot denoising framework tailored for low-angle CPWC ac…

AeroLite-MDNet: Lightweight Multi-task Deviation Detection Network for UAV Landing
Haiping Yang, Huaxing Liu, Wei Wu, Zuohui Chen, Ning Wu
https://arxiv.org/abs/2506.21635

AeroLite-MDNet: Lightweight Multi-task Deviation Detection Network for UAV Landing
Unmanned aerial vehicles (UAVs) are increasingly employed in diverse applications such as land surveying, material transport, and environmental monitoring. Following missions like data collection or inspection, UAVs must land safely at docking stations for storage or recharging, which is an essential requirement for ensuring operational continuity. However, accurate landing remains challenging due to factors like GPS signal interference. To address this issue, we propose a deviation warning sys…

FinStat2SQL: A Text2SQL Pipeline for Financial Statement Analysis
Quang Hung Nguyen, Phuong Anh Trinh, Phan Quoc Hung Mai, Tuan Phong Trinh
https://arxiv.org/abs/2506.23273

FinStat2SQL: A Text2SQL Pipeline for Financial Statement Analysis
Despite the advancements of large language models, text2sql still faces many challenges, particularly with complex and domain-specific queries. In finance, database designs and financial reporting layouts vary widely between financial entities and countries, making text2sql even more challenging. We present FinStat2SQL, a lightweight text2sql pipeline enabling natural language queries over financial statements. Tailored to local standards like VAS, it combines large and small language models in…

From Large-scale Audio Tagging to Real-Time Explainable Emergency Vehicle Sirens Detection
Stefano Giacomelli, Marco Giordano, Claudia Rinaldi, Fabio Graziosi
https://arxiv.org/abs/2506.23437

From Large-scale Audio Tagging to Real-Time Explainable Emergency Vehicle Sirens Detection
Accurate recognition of Emergency Vehicle (EV) sirens is critical for the integration of intelligent transportation systems, smart city monitoring systems, and autonomous driving technologies. Modern automatic solutions are limited by the lack of large scale, curated datasets and by the computational demands of state of the art sound event detection models. This work introduces E2PANNs (Efficient Emergency Pre trained Audio Neural Networks), a lightweight Convolutional Neural Network architectu…

Revisiting R: Statistical Envelope Analysis for Lightweight RF Modulation Classification
Srinivas Rahul Sapireddy, Mostafizur Rahman
https://arxiv.org/abs/2506.19956

Revisiting R: Statistical Envelope Analysis for Lightweight RF Modulation Classification
Modulation classification plays a crucial role in wireless communication systems, enabling applications such as cognitive radio, spectrum monitoring, and electronic warfare. Conventional techniques often involve deep learning or complex feature extraction, which, while effective, require substantial computational resources and memory. An early approach by Chan and Gadbois in 1985 introduced a theoretical method for modulation classification using a mathematically derived parameter called R. The…

Building Lightweight Semantic Segmentation Models for Aerial Images Using Dual Relation Distillation
Minglong Li, Lianlei Shan, Weiqiang Wang, Ke Lv, Bin Luo, Si-Bao Chen
https://arxiv.org/abs/2506.20688

Building Lightweight Semantic Segmentation Models for Aerial Images Using Dual Relation Distillation
Recently, there have been significant improvements in the accuracy of CNN models for semantic segmentation. However, these models are often heavy and suffer from low inference speed, which limits their practical application. To address this issue, knowledge distillation has emerged as a promising approach to achieve a good trade-off between segmentation accuracy and efficiency. In this paper, we propose a novel dual relation distillation (DRD) technique that transfers both spatial and channel r…

Nice - Emile Donovan talks to a local (to Ōtautahi) tech legend, Dr Tim Bell (U Canterbury CompSci lecturer) about teaching kids about computers https://www.rnz.co.nz/national/programmes/nights/audio/20…

Remote Rendering for Virtual Reality: performance comparison of multimedia frameworks and protocols
Daniel Mej\'ias, Inhar Yeregui, Roberto Viola, Miguel Fern\'andez, Mario Montagud
https://arxiv.org/abs/2507.00623

Remote Rendering for Virtual Reality: performance comparison of multimedia frameworks and protocols
The increasing complexity of Extended Reality (XR) applications demands substantial processing power and high bandwidth communications, often unavailable on lightweight devices. Remote rendering consists of offloading processing tasks to a remote node with a powerful GPU, delivering the rendered content to the end device. The delivery is usually performed through popular streaming protocols such as Web Real-Time Communications (WebRTC), offering a data channel for interactions, or Dynamic Adapt…

'Beneath the surface of external dominance lies an “invisible” European strength: a long-standing presence in open-source software (OSS) and open standards that form the backbone of critical digital infrastructures. From Linux and Python to core internet protocols, Europe has made significant
contributions to open source initiatives.'
3/n

Hecto: Modular Sparse Experts for Adaptive and Interpretable Reasoning
Sanskar Pandey, Ruhaan Chopra, Saad Murtaza Bhat, Ark Abhyudaya
https://arxiv.org/abs/2506.22919

Hecto: Modular Sparse Experts for Adaptive and Interpretable Reasoning
Mixture-of-Experts (MoE) models enable conditional computation by routing inputs to specialized experts, but these experts rely on identical inductive biases, thus limiting representational diversity. This static computation pathway is inefficient for inputs that require different types of reasoning and limits specialization and interpretability. We propose Hecto, a lightweight MoE architecture that leverages architectural heterogeneity by combining a GRU expert for temporal reasoning and an FF…

Wow, this just crossed my feed: https://www.euro-stack.info/docs/EuroStack_2025.pdf It describes the new 'EuroStack' of software, focused on helping the EU achieve digital sovereignty and arrest digital colonisation by US

Earworm warning: I've had this song in my head/under my breath since I hear it again (recently). It's superb. https://yewtu.be/watch?v=IZeWPScnolo or

Whelp, it's a cracking day for dryin' (laundry) here in Ōtautahi!

Wow, Substack signup is painful.

Just about to install Ubuntu Studio on my younger boy's clapped out (munted) school BYOD Thinkpad (purchased refurbished 5 yrs ago) so I can use it for music recording. Maybe. If it works.
How's your Friday night going?

Holy cow, the vocal virtuosity on this is second to none. https://yewtu.be/watch?v=Ct65mJmHQnw or https://www.youtube.com…

Richard Murphy provides an excellent characterisation of where western society was at its best (for most people, but not everyoe), and how it fell into the increasingly horrible situation it's in now... and how we could return to a 'better way' - what he refers to as 'mixed markets'. https://yewtu.be/watch?v=lrQh0_B…

Stunning day here in Ōtautahi.

Ok, checking if skymoth.app will let me crosspost from Mastodon -> BlueSky...

I've just set up (again, didn't seem to 'take' last time) POSSE (https://indieweb.org/POSSE - Publish (on your) Own Site, Syndicate Elsewhere) for this Mastodon account to my 'stub' account on BlueSky - if I understand it correctly, it should syndicate my posts here (in 'my ow…

This is a superb & unexpected rendition of a great (imho) song: https://yewtu.be/watch?v=VUhyV4zrBgY or https://www.y…

Wow, cleaning out my excess 😬 browser tabs last night, I came across this gem: https://ia.net/topics/markdown-and-the-slow-fade-of-the-formatting-fetish - this is something most people need to see, especially those in government and education.

There's a mysterious 'carbon sink' in the hinterlands of the South Island... absorbing more carbon than expected... I'm following this research with interest. https://www.rnz.co.nz/news/national/564407…

About to go out to lunch at a new-to-me (and my lunching compadre) Filipino restaurant... a voyage of gastronomical discovery awaits! Hoping 🤞 it's a 'keeper'!

Got a new keyboard with a different layout... thought it would be a good idea. And yeah, I never used that number pad anyway... but damn, the change has slowed me down... give me a minute to develop some new muscle memory...

Reckon it'd be pretty metal if Canada stopped Trump at the boarder, sending him back when he tries to get to the G7 Summit. No convicted felons/fascist autocrats allowed.

Kinda cool to see the programme for the UN's open source week... https://www.un.org/digital-emerging-technologies/content/open-source-week-2025 but, of course, it's already almost entirely co-opted by corporate interests. They shit on ev…

Had a weird dream-filled night. Dreamt I was in a series of pubs, but I could never order a beer because none of them had lists of beers available, nor did the barkeepers ever get to me... also, it occurs to me that they might not have been pubs. 😬 🤔

And, with all airborne balls in a relatively stable state, I'm calling it: beer o'clock.

Dunno about the rest of you, but I've been wondering for quite a few years how long it would take for one (or more) of the very very large baskets into which many people, businesses, and organisations have put *all* their eggs to suffer a catastrophic failure. Because it will happen. And some people will be caught out, and a lot will be lost. Because: unaware users & late stage capitalism.

Wow. https://themarkup.org/pixel-hunt/2025/04/28/how-california-sent-residents-personal-health-data-to-linkedin My opinion of LinkedIn was already vanishingly low (one way to ensure I won't read something is to po…

Hello all - for anyone who's interested, we'll be having our June monthly #libre | Free and Open Source Software (#FOSS) catchup tomorrow (Tues) evening, at 20:00 NZST. We meet at

My wife is working from home, and I'm listening to her and her colleagues having an online meeting via MS Teams - everything is "Teams" this or "Teams" that. 🤢 🤮 Poor folks being forced by their *gov't funded institution* to contend with such rubbish, poorly conceived, elitist (overpriced) digital technology, created by such an unworthy & psychotic entity (the Microsoft Corporation).

Crikey, just (finally) nailed the 'family photo gallery in a frame' functionality for my mum (with failing memory, in an assisted living situation) across the world. Can control it from here in Aotearoa (via screen share or ssh) with the extraordinarily handy 'RaspberryPi Connect' functionality. The key technologies are:
- Firefox (in kiosk mode),
- Immich (

Folks in NZ are starting to realise the implications of the impending EOL of Microsoft Windows 10 for our education system. In its finite wisdom, our MinOfEd has effectively locked all schools into either the use of MS Windows or Google ChromeOS. Both have fundamental issues like endless license hassles, EOLing OSs, ending support for Window or ChromeOS on specific computer models, etc. And that doesn't even start to talk about digital colonisation & sovereignty (see

Also just been watching a kōtare out the window. I startled it by walking past the window and then saw it sitting on a nearby street light. It then flew down and ate a few things off the rain slick road... worms, I'm guessing. Gorgeous to see its blue and gold highlights flitting around.

Yikes - just had 5 boxes of beers delivered... I think we ordered one... 🤔 Gulp. 🙃 😎