{"ID":5438772,"CreatedAt":"2026-07-01T01:17:58.482524686Z","UpdatedAt":"2026-07-03T09:43:49.071287852Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2606.31421","arxiv_id":"2606.31421","title":"Temporal Preservation over Processing: Diagnosing and Designing Spatiotemporal Single-Stage Video Detectors","abstract":"Single-stage video object detectors are increasingly deployed in time-critical applications, yet it remains unclear whether these models genuinely reason over temporal context or merely exploit a single informative frame-a gap hidden by standard metrics, which reward correct predictions regardless of how they are reached. We address this from two complementary directions: first, we propose TemporalLens, a model-agnostic diagnostic framework probing temporal dependence through controlled perturbations, structured occlusions, temporal shuffling, redundancy injection, and resolution degradation, revealing whether a detector actually uses information across time. Applied to stacked-frame 2D detectors and our YOLO-3D architecture, it exposes behavioural differences invisible to mAP: stacked 2D models collapse when the target frame is removed, while spatiotemporal models recover predictions from earlier frames, a signature of real temporal reliance. Second, we detail YOLO-3D, a modular real-time spatiotemporal detector built on YOLOv8, and show that simply preserving temporal depth through the backbone is the dominant performance driver (+3.7 pp mAP@50 at 32 frames averaged across scales). Together, the diagnostics and architecture turn \"does this detector reason over time?\" into a measurable, actionable question.","short_abstract":"Single-stage video object detectors are increasingly deployed in time-critical applications, yet it remains unclear whether these models genuinely reason over temporal context or merely exploit a single informative frame-a gap hidden by standard metrics, which reward correct predictions regardless of how they are reach...","url_abs":"https://arxiv.org/abs/2606.31421","url_pdf":"https://arxiv.org/pdf/2606.31421v1","authors":"[\"Karam Tomotaki-Dawoud\",\"Anna Hilsmann\",\"Peter Eisert\",\"Sebastian Bosse\"]","published":"2026-06-30T09:44:21Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.AI\"]","methods":"[]","has_code":false}