Ad-hoc video search
태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.
- 하위 태스크
- 0
- 평가표
- 5
- 보존된 평가 기록
- 15
The Ad-hoc search task ended a 3 year cycle from 2016-2018 with a goal to model the end user search use-case, who is searching (using textual sentence queries) for segments of video containing persons, objects, activities, locations, etc. and combinations of the former. While the Internet Archive (IACC.3) dataset was adopted between 2016 to 2018, starting in 2019 a new data collection based on Vimeo Creative Commons (V3C) will be adopted to support the task for at least 3 more years.
Given the test collection (V3C1 or IACC.3), master shot boundary reference, and set of Ad-hoc queries (approx. 30 queries) released by NIST, return for each query a list of at most 1000 shot IDs from the test collection ranked according to their likelihood of containing the target query.
논문
이 태스크와 하위 분야의 관련 논문입니다. 실제 성능 비교는 평가 결과에서 확인하세요.
Improving Interpretable Embeddings for Ad-hoc Video Search with Generative Captions and Multi-word Concept Bank
Interpretable Embedding for Ad-hoc Video Search
An overview on the evaluated video retrieval tasks at TRECVID 2022
Renmin University of China at TRECVID 2022: Improving Video Search by Feature Fusion and Negation Understanding
(Un)likelihood Training for Interpretable Embedding
Lightweight Attentional Feature Fusion: A New Baseline for Text-to-Video Retrieval
TRECVID 2020: A comprehensive campaign for evaluating video retrieval tasks across multiple application domains
SEA: Sentence Encoder Assembly for Video Retrieval by Textual Queries
TRECVID 2019: An Evaluation Campaign to Benchmark Video Activity Detection, Video Captioning and Matching, and Video Search & Retrieval
Dual Encoding for Video Retrieval by Text
W2VV++: Fully Deep Learning for Ad-hoc Video Search
Dual Encoding for Zero-Example Video Retrieval
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.