Abdelrrahman Moubane — WritingWriting on retrieval, computer vision, and applied AI.https://abdelpy.github.io/Why Binary Change Detection Isn't Enoughhttps://abdelpy.github.io/writing/binary-vs-multi-label-change-detection/https://abdelpy.github.io/writing/binary-vs-multi-label-change-detection/A practical look at what disappears when every temporal difference is reduced to changed or unchanged.Fri, 19 Jun 2026 00:00:00 GMTInside FOTBCD-CC: analysing 321,200 aerial captionshttps://abdelpy.github.io/writing/fotbcd-cc-dataset-analysis/https://abdelpy.github.io/writing/fotbcd-cc-dataset-analysis/An analysis of 321,200 model-generated aerial captions: dataset composition, descriptions per location, language patterns, and candidates for visual review.Thu, 17 Sep 2026 00:00:00 GMTDistilling Qwen3.8-Max into a smaller aerial change modelhttps://abdelpy.github.io/writing/fotbcd-cc-fine-tuning/https://abdelpy.github.io/writing/fotbcd-cc-fine-tuning/The decisions behind making aerial change captioning affordable: teacher supervision, prompt caching, LoRA adaptation, and quality evaluation.Wed, 16 Sep 2026 00:00:00 GMTTraining a multimodal model to retrieve aerial changehttps://abdelpy.github.io/writing/fotbcd-cc-multimodal-training/https://abdelpy.github.io/writing/fotbcd-cc-multimodal-training/The design decisions behind teaching DINOv3 and Qwen to connect descriptions of change with before-and-after aerial imagery.Wed, 16 Sep 2026 00:00:00 GMTZero-Shot Damage Assessment: Evaluating MBCTD on the 2025 LA Wildfireshttps://abdelpy.github.io/writing/la-wildfire-damage-assessment/https://abdelpy.github.io/writing/la-wildfire-damage-assessment/Testing whether a building-change model trained on ordinary urban development can identify wildfire destruction in unseen Los Angeles imagery.Fri, 12 Jun 2026 00:00:00 GMT