Đorđe Klisura, Astrid R Bernaga Torres, Anna Karen Gárate-Escamilla, Rajesh Roshan Biswal, Ke Yang, Hilal Pataci, Anthony Rios. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.

Globally, cyberattacks are growing and mutating each month. Intelligent Intrusion Network Detection Systems are developed to analyze and detect anomalous traffic to face these threats. A way to address this is by using network flows, an aggregated version of communications between devices. Network Flow datasets are used to train Artificial Intelligence (AI) models to classify specific attacks. Training these models requires threat samples usually generated synthetically in labs as capturing them on operational network is a challenging task. As threats are fast-evolving, new network flows are continuously developed and shared. However, using old datasets is still a popular procedure when testing models, hindering a more comprehensive characterization of the advantages and opportunities of recent solutions on new attacks. Moreover, a standardized benchmark is missing rendering a poor comparison between the models produced by algorithms. To address these gaps, we present a benchmark with fourteen recent and preprocessed datasets and study seven categories of algorithms for Network Intrusion Detection based on Network Flows. We provide a centralized source of pre-processed datasets to researchers for easy download. All dataset are also provided with a train, validation and test split to allow a straightforward and fair comparison between existing and new solutions. We selected open state-of-the-art publicly available algorithms, representatives of diverse approaches. We carried out an experimental comparison using the Macro F1 score of these algorithms. Our results highlight each model operation on dataset scenarios and provide guidance on competitive solutions. Finally, we discuss the main characteristics of the models and benchmarks, focusing on practical implications and recommendations for practitioners and researchers.

We propose an end-to-end solution for recognizing fingerspelling using multi-scale attention with fixed-queries. Fingerspelling recognition in the wild gets challenging because of the multiple sub-problems involved - detecting the signing hand, tracking it across frames, and recognizing subtle variations in a hand gesture. While the current state-of-the-art handles these with external face/hand detectors, optical flow features, and iteratively refining the attention maps, our work proposes a deep learning model that takes in the RGB videos and recognizes fingerspelling with a single forward pass. Without any frame-level supervision, our proposed model learns to pay attention to informative regions in each frame, such as fingers, hand, and face, to recognize signs. Multi-scale features from these attended regions are then processed using a recurrent neural network to recognize the alphabet sequentially. We train our model using a curriculum learning strategy with simpler samples at the beginning, followed by challenging samples at a later stage. We have evaluated our approach on Chicago Fingerspelling Wild and WildPlus datasets and have achieved about 8% and 4% improvements, respectively, compared to the current state-of-the-art methods. Further analysis of our method shows that our attention mechanism is intuitive from a human perspective, and visualizing it offers useful insights into the working of the model.

While deep learning based pedestrian detection systems have continued to scale new heights in recent times, the performance of such algorithms tends to degrade under challenging illumination conditions. This causes a bottleneck in ready portability of such systems to Advanced Driver Assistance Systems (ADAS), where consistent performance across varying environmental lighting is desired. Inspired by the concept of dark knowledge, this paper proposes a novel guided deep network that distills knowledge from a multi-modal pedestrian detector. The proposed network learns to extract both RGB and thermal-like features from RGB images alone, thus compensating for the requirement of significantly costly automotive-grade thermal cameras. Compelling detection performance in severe lighting conditions is demonstrated on a publicly available night-time pedestrian dataset — KAIST. We achieve an effective miss-rate of 12% lower than the recent state-of-the-art methods.