"Hope" at NakbaArchiveClassifier Shared Task: Transfer Learning-Based CNN Models for Infrastructure Damage Detection

Team Hope's ResNet34 model, fine-tuned with ImageNet pretrained weights for 25 epochs using the Adam optimizer and a learning rate of 1e-4, achieved 81% accuracy on the NakbaArchiveClassifier Shared Task at Nakba-NLP 2026, outperforming other tested architectures including ResNet50 and EfficientNet-B0. The system, developed by Lojien AlKhidir and HebaTalla Abdelhady, classified 2,001 Instagram images into destruction and not_destruction categories, demonstrating that moderate-depth CNNs can generalize effectively in low-resource visual classification tasks.

Abstract This paper describes Team Hope’s system for the NakbaArchiveClassifier Shared Task at Nakba-NLP 2026. The task focuses on binary classification of social media images into two categories: destruction and not destruction. We evaluated multiple convolutional neural network architectures using transfer learning, including ResNet34, ResNet50, EfficientNet-B0, and a fine-tuned ResNet34 variant with staged training. All models were initialized with ImageNet pretrained weights and fine-tuned on the provided dataset of 2,001 images. The dataset is moderately imbalanced and contains visually diverse Instagram images depicting intact and damaged infrastructure. Our best-performing model, ResNet34 trained for 25 epochs with Adam optimizer and a learning rate of 1e-4, achieved 81% accuracy on the evaluation platform. We provide a comparative analysis of the tested architectures and discuss the impact of model depth, training duration, and class imbalance. Given the political and ethical sensitivity of the dataset, we also include a discussion of responsible AI considerations and potential limitations. Our findings suggest that moderate-depth architectures can generalize effectively in low-resource, contextually complex visual classification tasks.- Anthology ID: - 2026.nakbanlp-1.25 - Volume: Proceedings of the 2nd International Workshop on Nakba Narratives as Language Resources @ LREC 2026 /volumes/2026.nakbanlp-1/ - Month: - May - Year: - 2026 - Address: - Palma, Mallorca Spain - Editors: Mustafa Jarrar /people/mustafa-jarrar/ , Mo El-Haj /people/mo-el-haj/ , Amal Haddad /people/amal-haddad/unverified/ , Serin Atiani /people/serin-atiani/unverified/ , Shadi Abudalfa /people/shadi-abudalfa/ , Terry Regier /people/terry-regier/unverified/ , Paul Rayson /people/paul-rayson/ , Khalil Sima’an /people/khalil-simaan/ , Camille Mansour /people/camille-mansour/unverified/ - Venues: NakbaNLP /venues/nakbanlp/ | WS /venues/ws/ - SIG: - Publisher: - ELRA Language Resources Association ELRA - Note: - Pages: - 187–190 - Language: - External URL: https://lrec.elra.info/lrec2026-ws-nakbanlp-25 https://lrec.elra.info/lrec2026-ws-nakbanlp-25 - DOI: 10.63317/233j9kgmifw4 https://doi.org/10.63317/233j9kgmifw4 - Cite ACL : - Lojien AlKhidir and HebaTalla Abdelhady. 2026. "Hope" at NakbaArchiveClassifier Shared Task: Transfer Learning-Based CNN Models for Infrastructure Damage Detection https://aclanthology.org/2026.nakbanlp-1.25/ . In Proceedings of the 2nd International Workshop on Nakba Narratives as Language Resources @ LREC 2026 , pages 187–190, Palma, Mallorca Spain . ELRA Language Resources Association ELRA . - Cite Informal : “Hope” at NakbaArchiveClassifier Shared Task: Transfer Learning-Based CNN Models for Infrastructure Damage Detection https://aclanthology.org/2026.nakbanlp-1.25/ AlKhidir & Abdelhady, NakbaNLP 2026