Abstract:Transformer-based detectors have advanced small-object detection, but they often remain inefficient and vulnerable to background-induced query noise, which motivates deep decoders to refine low-quality queries. We present HELP (Heatmap-guided Embedding Learning Paradigm), a noise-aware positional-semantic fusion framework that studies where to embed positional information by selectively preserving positional encodings in foreground-salient regions while suppressing background clutter. Within HELP, we introduce Heatmap-guided Positional Embedding (HPE) as the core embedding mechanism and visualize it with a heatbar for interpretable diagnosis and fine-tuning. HPE is integrated into both the encoder and decoder: it guides noise-suppressed feature encoding by injecting heatmap-aware positional encoding, and it enables high-quality query retrieval by filtering background-dominant embeddings via a gradient-based mask filter before decoding. To address feature sparsity in complex small targets, we integrate Linear-Snake Convolution to enrich retrieval-relevant representations. The gradient-based heatmap supervision is used during training only, incurring no additional gradient computation at inference. As a result, our design reduces decoder layers from eight to three and achieves a 59.4% parameter reduction (66.3M vs. 163M) while maintaining consistent accuracy gains under a reduced compute budget across benchmarks. Code Repository: this https URL
| Comments: | Accepted to ACM ICMR 2026; 14 pages, 6 figures, and 4 tables |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2604.15065 [cs.CV] |
| (or arXiv:2604.15065v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2604.15065 arXiv-issued DOI via DataCite (pending registration) |
From: Kangning Cui [view email]
[v1]
Thu, 16 Apr 2026 14:28:36 UTC (4,537 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。