VLSP 2021 Evaluation Campaign

The VLSP 2021 evaluation campaign deals with 8 shared-tasks for text and speech processing.

Text processing

  • Named Entity Recognition: Recognizing named entities in documents (datasets released by VNU-HUS)
    • Organizers: Trần Mai Vũ (VNU-UET), Nguyễn Thị Minh Huyền (VNU-HUS), Hà Mỹ Linh (VNU-HUS)
  • Machine Translation: English - Vietnamese and/or Chinese - Vietnamese machine translation (datasets released by the Association for VLSP and VNU-UET)
    • Organizers: Nguyễn Văn Vinh (VNU-UET), Trần Hồng Việt (UNETI), Nguyễn Lê Minh (JAIST) 
  • Vietnamese Machine Reading Comprehension: Extraction-based machine reading comprehension on Vietnamese Wikipedia-based texts (corpus released by VNUHCM-UIT)
    • Organizers: Nguyễn Lưu Thuỳ Ngân (VNUHCM-UIT), Nguyễn Văn Kiệt (VNUHCM-UIT), Lưu Thanh Sơn (VNUHCM-UIT), Huỳnh Văn Tín (VNUHCM-UIT), Nguyễn Thành Luân (VNUHCM-UIT), Trần Quốc Sơn (Denison University, USA)
  • Vietnamese and English-Vietnamese Textual Entailment: Recognizing textual entailment relation between 2 sentences (corpus released by VNU-HUS, funded by a VINIF project)
    • Organizers: Hoàng Tuấn Anh (Viettravel), Nguyễn Thị Minh Huyền (VNU-HUS), Nguyễn Lưu Thuỳ Ngân (VNUHCM-UIT), Trần Thị Oanh (VNU-IS), Ngô Thế Quyền (VNU-HUS)
  • Image Captioning: vieCap4H Challenge: Automatic image caption generation for healthcare domains in Vietnamese (datasets released by VNU-HUS, funded by a VINIF project)
    • Organizers: Le Minh Thao (Deakin University, Australia), Hoàng Tuấn Anh (Viettravel), Đặng Hoàng Long (Deakin University, Australia), Nguyen Thanh Sơn (A*STAR, Singapore), Vũ Xuân Sơn (Umeå University, Sweden)

Speech processing

  • Automatic Speech Recognition for Vietnamese: Automatic speech recognition for conversational speech (datasets released by Association for VLSP)
    • Organizers: Đỗ Văn Hải (TLU)
  • Vietnamese Text-To-Speech: Speech synthesis from spontaneous speech (datasets released by Vbee)
    • Organizers: Nguyễn Thị Thu Trang (HUST), Nguyễn Hoàng Kỳ (Vbee), Phạm Quang Minh (Vbee)
  • Vietnamese Speaker Verification: Open and Closed Challenge (datasets released by HUST)
    • Organizers: Vi Thành Đạt (VNG), Phạm Việt Thành (HUST), Nguyễn Thị Thu Trang (HUST)

VLSP shared-tasks aim at promoting the most efficient methods for these important tools. The organization of these campaigns with sponsorships from academia and industry permit to build and offer to the VLSP community gold datasets for training and testing Vietnamese text and speech processing systems.

Participants of all speech shared tasks this year have to contribute or join to build the dataset before receiving it. The main task is to transcribe or to correct the transcription or to verify the same identity for a small part of the dataset.

The participants to the evaluation campaign will be asked to present their system in a dedicated paper.

Shared-task registration

Please fill in this form to participate in one or more VLSP 2021 shared-tasks (open on Aug 5, 2021).