arXiv:2609.22500v1 Announce Type: new Abstract: Autonomous navigation requires precise and efficient semantic segmentation, yet existing frame-based approaches remain limited by motion blur, glare, l...
arXiv:2610.09710v1 Announce Type: new Abstract: ANN-to-SNN conversion offers a practical route toward energy-efficient spiking Vision-Language-Action (VLA) models by bypassing the substantial cost of...
REACT is a fully spiking state‑space model that processes raw event‑camera data one event at a time, avoiding temporal accumulation and its associated delay. It employs a complex‑valued spiking neuron (C‑SiLIF) whose dynamics are driven by the inter‑event interval, enabling continuous‑time state updates at microsecond resolution. Evaluated on gesture recognition and time‑to‑collision estimation, REACT achieves low latency (4.6 ms) and high accuracy, supports anytime prediction, zero‑shot transfer, and INT8 quantization, dramatically reducing energy consumption.
SV-TAD introduces native sparse 2D convolutions to enable convolutional adapters to operate directly on dynamically pruned token sets, addressing the inefficiency of dense reconstruction in token‑selection methods. By integrating this primitive into a temporal action detection framework, the approach cuts VideoMAEv2‑L computation by up to 64% and achieves 2.2× faster inference while preserving state‑of‑the‑art accuracy on THUMOS‑14 and ActivityNet‑1.3. When applied to InternVideoNext‑L, it surpasses prior state‑of‑the‑art performance at roughly half the computational cost and naturally supports auxiliary task tokens for improved fine‑grained assembly detection on ATTACH.
SV‑TAD introduces native sparse 2D convolutions to enable convolutional adapters to operate directly on dynamically pruned token sets, addressing the inefficiency of existing adapters that do not reduce inference‑time computation. By integrating this primitive, SV‑TAD cuts VideoMAEv2‑L computation by up to 64% and achieves 2.2× faster inference while preserving state‑of‑the‑art accuracy on THUMOS‑14 and ActivityNet‑1.3. When applied to InternVideoNext‑L, the approach surpasses prior state‑of‑the‑art performance at roughly half the computational cost and also supports auxiliary task tokens, improving fine‑grained assembly detection on ATTACH.
The paper introduces four new event‑based vision datasets created with the ANTShapes simulation tool, designed to support object classification research using spiking neural networks (SNNs). These datasets vary in difficulty and are benchmarked against established spiking datasets such as N‑MNIST, CIFAR10‑DVS, DVSGesture, and POKER‑DVS using a convolutional SNN. The work provides detailed, high‑quality datasets for future experiments and validates ANTShapes as a suitable tool for generating event‑based vision data.
arXiv:2605. 05895v2 Announce Type: replace-cross Abstract: Modern AI-generated videos are photorealistic at the single-frame level, leaving inter-frame dynamics as the main remaining axis for detection.
arXiv:2608. 19238v1 Announce Type: cross Abstract: Spiking Transformers model token interactions primarily through spiking self-attention (SSA).
arXiv:2610.00350v1 Announce Type: new Abstract: Spiking Neural Networks (SNNs) offer an energy-efficient approach to processing event-camera data, yet out-of-distribution (OOD) detection remains chal...
The paper presents a lightweight, training‑free framework for real‑time unsupervised object discovery from asynchronous event camera streams. It introduces a linear‑time Spatio‑Temporal Probabilistic Event Filter (SPEF) that adaptively distinguishes salient motion from noise, and an Event Morton Code Clustering (EMCC) module that efficiently groups events without costly distance calculations. Experiments on E‑MLB, FRED, and eTraM datasets show SPEF outperforms classical filters and competes with learning‑based methods, while EMCC achieves the highest accuracy and fastest execution among density‑based clustering baselines.
The paper introduces a confidence‑normalized continuous multi‑timescale representation for event cameras, using logarithmic B‑spline temporal encoding and a geometry‑aware local confidence mechanism. When paired with a fixed feed‑forward EventCenterNet detector, this representation outperforms the compact CSTR representation on the PEDRo and Gen1 datasets. Additionally, a recursive exponential‑polynomial approximation is proposed to allow efficient event‑by‑event updates while maintaining detection performance.
NeuroDETR is a novel multiscale discrete spatiotemporal tokenization framework designed for asynchronous EEG event detection. It combines structured tokenization, event-aware feature refinement, and a sparse attention module called FOCAL to capture both brief and sustained neural responses while preserving spatial hierarchy across EEG channels. Experiments on eight diverse EEG datasets show that NeuroDETR achieves state‑of‑the‑art performance in jointly classifying events and estimating their onsets and offsets.