activity
20242026
most citedR2T: Rule-Encoded Loss Functions for Low-Resource Sequence Tagging

1 citations · 2 across the 9 of their papers we have counts for

collaborators

9 papers

cs.ET2026

OmniEye: Efficient Multimodal Forensic Video Intelligence for Law-Enforcement Body-Worn Cameras

Mamadou K. Keita, Angela Srbinovska, Anita Srbinovska +12

We introduce OmniEye, a multimodal video intelligence system for law-enforcement training and review (source code available on request to verified law-enforcement and public-safety…

cs.CR2026

Computer Science Conferences Should Require Nonrepudiable Experimental Results

Mamadou K. Keita, Christopher Homan

This position paper argues that computer science conferences should require tamper-evident, nonrepudiable attestations of experimental results. We name the underlying problem exper…

cs.CL2026

Where Are We At with Automatic Speech Recognition for the Bambara Language?

Seydou Diallo, Yacouba Diarra, Mamadou K. Keita +3

This paper introduces the first standardized benchmark for evaluating Automatic Speech Recognition (ASR) in the Bambara language, utilizing one hour of professionally recorded Mali…

cs.LG2025

InstructLR: A Scalable Approach to Create Instruction Dataset for Under-Resourced Languages

Mamadou K. Keita, Sebastien Diarra, Christopher Homan +1

Effective text generation and chat interfaces for low-resource languages (LRLs) remain a challenge for state-of-the-art large language models (LLMs) to support. This is mainly due…

cs.LG2025

NSL-MT: Linguistically Informed Negative Samples for Efficient Machine Translation in Low-Resource Languages

Mamadou K. Keita, Christopher Homan, Huy Le

We introduce negative space learning machine translation (NSL-MT), a training method for underresourced languages, that augments limited parallel data with synthetically generated…

cs.CL2025★ 1 cited

R2T: Rule-Encoded Loss Functions for Low-Resource Sequence Tagging

Mamadou K. Keita, Christopher Homan, Sebastien Diarra

We introduce the Rule-to-Tag (R2T) framework, a hybrid approach that integrates a multi-tiered system of linguistic rules directly into a neural network's training objective. R2T's…