A Unified Framework for scalable Vehicle Trajectory Prediction, ECCV 2024
Official code, open-set data splits, and reproduction scripts for our ICDAR WML 2026 paper: "Cross-Dataset Transfer in Open-Set Writer Identification".
Unified multiclass and cross-corpus evaluation of UNSW-NB15 and CICIDS-2017 (ICEEICT 2027, under review)
Benchmarking ResNet50 and EfficientNetB0 for cross-dataset visual concrete-defect classification, with group-aware five-fold validation, transfer evaluation, reproducible fold assignments, and reported metrics.
Fine-tuned DistilBERT for political misinformation classification with cross-dataset evaluation across LIAR and Fakeddit benchmarks, compared against Logistic Regression, Naive Bayes, and SVM baselines.
To associate your repository with the cross-dataset-evaluation topic, visit your repo's landing page and select "manage topics."