Progress in machine understanding of sign languages has been slow and hampered by limited data. In this paper, we present FSboard, an American Sign Language finger-spelling dataset situated in a mobile text entry use case, collected from 147 paid and consenting Deaf signers using Pixel 4A selfie cameras in a variety of environments. Fin-gerspelling recognition is an incomplete solution that comprises only a small part of sign language translation, but it could provide some immediate benefit to Deaf/Hard of Hearing signers while more broadly capable technology develops. At >3 million characters in length and >250 hours in duration, FSboard is the largest fingerspelling recognition dataset to date by a factor of >10x. As a simple baseline, we finetune 30 Hz MediaPipe Holistic landmark inputs into ByT5-Small and achieve 11.1% Character Error Rate (CER) on a test set with unique phrases and signers. This quality degrades gracefully when decreasing frame rate and excluding face/body landmarks—plausible optimizations to help with on-device performance—but falls short of human performance measured at 2.2% CER.1

Machine learning for sign languages is bottlenecked by data. In this paper, we present YouTube-ASL, a large-scale, open-domain corpus of American Sign Language (ASL) videos and accompanying English captions drawn from YouTube. With ~1000 hours of videos and >2500 unique signers, YouTube-ASL is ~3x as large and has ~10x as many unique signers as the largest prior ASL dataset. We train baseline models for ASL to English translation on YouTube-ASL and evaluate them on How2Sign, where we achieve a new fine-tuned state of the art of 12.397 BLEU and, for the first time, nontrivial zero-shot results.

Modeling the dynamics of heart and lung tissue is challenging because the tissue deforms between data acquisitions. To reconstruct complete volumes, sample data captured at different times and locations must be combined. This paper presents a novel end-to-end, data driven framework for the complete reconstruction of deforming tissue volumes. This framework is a joint optimization over an undeformed tissue volume, a deformation map that describes tissue motion forgiven pose parameters (i.e. breathing and heartbeat), and an estimate of those parameters for each data acquisition. Tissue motion is modeled by deforming a reference volume with a cubic B-spline free form deformation, and we use Isomap to derive initial estimates of the pose of sample data. An iterative method is used to simultaneously solve for the reference volume and deformation map while updating the pose estimates. This same process is demonstrated on 4D CT lung data and heart/lung MR data.

Fair Queuing (FQ) algorithms provide isolation between packet flows, allowing max-min fair sharing of a link even when flows misbehave. However, fairness comes at the expense of per-flow state. To keep the memory requirement independent of the flow count, the router can isolate aggregates of flows, rather than individual flows. We investigate the feasibility of protecting individual flows under such aggregate isolation in the context of Multiple Queue Fair Queuing (MQFQ), where the router maintains a fixed number of queues and allows each flow to access multiple queues. MQFQ places packets into the shortest queue associated with their flow. The extra queues protect the flow against congestion caused by a misbehaving flow in a shared queue. However, multiple per-flow queues also enable the misbehaving flow to increase its unfairly acquired fraction of the link capacity. We discuss avoidance of packet reordering within a flow and compare MQFQ with prior schemes for aggregate scheduling.