mlsanityData to 14 Oct 2026

DERA: Detached Edge-Residual Adaptation for Prohibited item Detection

arXiv 2609.12411 ↗First posted 11 Sep 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Prohibited-item detection in X-ray imagery remains challenging due to object superposition, weak texture, and material clutter which obscure both semantic appearance and object boundaries. We propose DERA, a Detached Edge-Residual Adaptation framework for prohibited item detection under X-ray imagery. DERA combines hierarchical visual features with a parallel pixel-difference edge pyramid and learns an object-specific boundary prior from training-time contours of the instance masks. The detached prior gates edge-sensitive features, which are injected into the early visual stages through residual heads. This staged design preserves the foundation detector at the start of adaptation, isolates boundary supervision from semantic feature learning, and restricts the final adaptation stage to only (14.7)K trainable parameters. Evaluated on PIDray, CLCXray, and STCray, DERA improves the baseline by 3.1, 1.6, and 2.4 AP points, respectively.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 38 of its 51 references that are in the map. First posted on arXiv 11 Sep 2026.

Builds on

38 of 51 references.

References · 38

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
  3. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,639
  4. 2020End-to-End Object Detection with Transformerscited by 7,414
  5. 2020Deformable DETR: Deformable Transformers for End-to-End Object Detectioncited by 3,422
  6. 2023Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detectioncited by 3,332
  7. 2024SAM 2: Segment Anything in Images and Videoscited by 2,863
  8. 2022Visual Prompt Tuningcited by 1,681
  9. 2022DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detectioncited by 1,319
  10. 2020Gradient Surgery for Multi-Task Learningcited by 1,256
  11. 2021Grounded Language-Image Pre-trainingcited by 1,156
  12. 2017GradNorm: Gradient Normalization for Adaptive Loss Balancing in Deep Multitask Networkscited by 834
26 more
  1. 2021MDETR -- Modulated Detection for End-to-End Multi-Modal Understandingcited by 819
  2. 2022AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognitioncited by 719
  3. 2021RegionCLIP: Region-based Language-Image Pretrainingcited by 616
  4. 2015Holistically-Nested Edge Detectioncited by 592
  5. 2017Learning multiple visual domains with residual adapterscited by 557
  6. 2021Conflict-Averse Gradient Descent for Multi-task Learningcited by 459
  7. 2019Which Tasks Should Be Learned Together in Multi-task Learning?cited by 375
  8. 2021Pixel Difference Networks for Efficient Edge Detectioncited by 168
  9. 2019Gated-SCNN: Gated Shape CNNs for Semantic Segmentationcited by 166
  10. 2016Richer Convolutional Features for Edge Detectioncited by 143
  11. 2022PIDNet: A Real-time Semantic Segmentation Network Inspired by PID Controllerscited by 84
  12. 2019Bi-Directional Cascade Network for Perceptual Edge Detectioncited by 72
  13. 2019SIXray : A Large-scale Security Inspection X-ray Benchmark for Prohibited Item Discovery in Overlapping Imagescited by 52
  14. 2020Occluded Prohibited Items Detection: an X-ray Security Inspection Benchmark and De-occlusion Attention Modulecited by 39
  15. 2021Towards Real-World Prohibited Item Detection: A Large-Scale X-ray Benchmarkcited by 31
  16. 2021Towards Real-world X-ray Security Inspection: A High-Quality Benchmark and Lateral Inhibition Module for Prohibited Items Detectioncited by 30
  17. 2024Time-, Memory- and Parameter-Efficient Visual Adaptationcited by 28
  18. 2020Towards Automatic Threat Detection: A Survey of Advances of Deep Learning within X-ray Security Imagingcited by 25
  19. 2024BEFUnet: A Hybrid CNN-Transformer Architecture for Precise Medical Image Segmentationcited by 11
  20. 2024AO-DETR: Anti-Overlapping DETR for X-Ray Prohibited Items Detectioncited by 9
  21. 2025STING-BEE: Towards Vision-Language Model for Real-World X-ray Baggage Security Inspectioncited by 5
  22. 2024Dual-view X-ray Detection: Can AI Detect Prohibited Items from Dual-view X-ray Images like Humans?cited by 4
  23. 2025CSPCL: Category Semantic Prior Contrastive Learning for Deformable DETR-Based Prohibited Item Detectorscited by 3
  24. 2025Superpowering Open-Vocabulary Object Detectors for X-ray Visioncited by 3
  25. 2025Can a Second-View Image Be a Language? Geometric and Semantic Cross-Modal Reasoning for X-ray Prohibited Item Detectioncited by 1
  26. 2026Falcon: Functional Assembly and Language for Compositional Reasoning in X-raycited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025ALDI-ray: Adapting the ALDI Framework for Security X-ray Object Detection
  2. 2026OneFocus: Enabling Real-World X-ray Security Screening with a Unified Vision-Language Model
  3. 2025Taming Generative Synthetic Data for X-ray Prohibited Item Detection
  4. 2026Localize Any Object in X-Ray Security Scans without Human Annotation
  5. 2024Open-Vocabulary X-ray Prohibited Item Detection via Fine-tuning CLIP
  6. 2026GSA-YOLO: A High-Efficiency Framework via Structured Sparsity and Adaptive Knowledge Distillation for Real-Time X-ray Security Inspection
  7. 2024Weak-to-Strong 3D Object Detection with X-Ray Distillation
  8. 2025Leveraging Confident Image Regions for Source-Free Domain-Adaptive Object Detection