2022 Open Challenges in Continuous Sign Language Recognition Workshop at ECCV

9:00-9:05

Opening (5min)

9:05-9:15

Challenge presentation (10min): ECCV 2022 Sign Spotting Challenge: Dataset, Design and Results - Manuel Vazquez-Enriquez (Vigo University), Jose Alba-Castro (Vigo University), Laura Docio-Fernandez (Vigo University), Julio C. S. Jacques Junior (Computer Vision Center), Sergio Escalera (University of Barcelona / Computer Vision Center)

9:15-9:30

Winner track 1 (10min) + Q&A (5min): Hierarchical I3D for Sign Spotting - Ryan Wong (University of Surrey), Necati Cihan Camgoz (University of Surrey), Richard Bowden (University of Surrey)

9:30-9:45

Winner track 2 (10min) + Q&A (5min): Multi-modal Sign Language Spotting by Multi/One-shot Learning - Landong Liu (University of Science and Technology of China), Wengang Zhou (University of Science and Technology of China), Weichao Zhao (University of Science and Technology of China), Hezhen Hu (University of Science and Technology of China), Houqiang Li (University of Science and Technology of China)

9:45-10:00

Workshop paper (10min ) + Q&A (5min): Sign Spotting via Multi-modal Fusion and Testing Time Transferring - Hongyu Fu (Tsinghua University), Chen Liu (University of Technology Sydney), Xingqun Qi (University of Technology Sydney), Beibei Lin (National University of Singapore), Lincheng Li (NetEase Fuxi AI Lab), Li Zhang (Tsinghua University), Xin Yu (University of Technology Sydney)

10:00-10:40

Annelies Braffort - "What is a sign? A multidisciplinary overview for a finer definition for automatic processing" (30min + 10min for Q&A)

Short description: Most works on the automatic processing of sign languages manipulate the notion of sign by considering it as the equivalent of a word/gloss in written languages. But what is it really? This keynote proposes to revisit the notion of sign by drawing on current knowledge in linguistics and science of movement, in order to identify the limits and possible progress in automatic processing, and more particularly in automatic recognition and sign spotting.

10:40-11:20

Richard Bowden - "Translating between Sign and Spoken Languages" (30min + 10min for Q&A)

Short description: This talk will discuss ongoing work across several linked projects into sign language translation. It will discuss the fundamentals of sign, the challenges of translation and recent work in translating sign language videos to spoken language as well as generating photo realistic video from spoken language sentences.

11:20-11:35

Break (15min)

11:35-12:15

Andrew Zisserman - "Learning  Sign Spotting from Signer-Interpreted TV" (30min + 10min for Q&A)

12:15-12:55

Lale Akarun - "Transfer learning for low-resource sign languages" (30min + 10min for Q&A)

12:55-13:00

Closing (5min)