Sign language recognition datasets
Comparison
id | Name | Country | Classes | Subjects | Samples | Data | Language level | Type | Annotations | Availability |
---|---|---|---|---|---|---|---|---|---|---|
1 | DGS Kinect 40 | Germany | 40 | 15 | 3000 | Word | Videos, multiple angles | Contact Author | ||
2 | RWTH-PHOENIX-Weather | Germany | 1200 | 9 | 45760 | 53gb | Sentence | Videos | Face, hand, end/start (unfinished) | Publicly Available |
3 | SIGNUM | Germany | 450 | 25 | 33210 | 920gb | Sentence | Videos | 1TB contact author | |
4 | GSL 20 | Greek | 20 | 6 | ~840 | Word | Contact Author | |||
5 | Boston ASL LVD | USA | 3300+ | 6 | 9800 | Word | Videos, multiple angles | hand,end/start | Publicly Available | |
6 | PSL Kinect 30 | Poland | 30 | 1 | 30×10=300 | ~1.2gb | Word | Videos, depth from Kinect camera | Publicly Available | |
7 | PSL ToF 84 | Poland | 84 | 1 | 84×20=1680 | ~33gb | Word | Videos, ToF camera | Publicly Available | |
8 | PSL 101 | Poland | ? | ? | ? | ? | ? | ? | Contact Author | |
9 | LSA64 | Argentina | 64 | 10 | 3200 | 20gb | Word | Videos(rgb) | Hands and Head position | Publicly Available |
10 | BosphorusSign | Turkey | Non available | |||||||
11 | MSR Gesture 3D | USA | 12 | 10 | 336 | 28mb | Word | Videos (depth) | Publicly Available | |
12 | DEVISIGN-G | China | 36 (letters/numbers) | 8 | 432 | ? | Word | videos (rgb) | Contact author | |
13 | DEVISIGN-D | China | 500 | 8 | 6000 | ? | Word | videos (rgb) | Contact author | |
14 | DEVISIGN-L | China | 2000 | 8 | 24000 | ? | Word | videos (rgb) | Contact author | |
15 | IIITA -ROBITA | India | 23 | ? | 284mb | Word | videos (rgb) 320x240 | Contact author | ||
16 | Purdue ASL | USA | ? | 14 (only 5 available) | ? | ? | Word/sentence | Videos, RGB | Request DVDs/HD | |
17 | CUNY ASL | USA | ? | 8 | ~33000 glosses, | ? | Sentence | videos (rgb), mocap data | Signstream | Unknown |
18 | SignsWorld Atlas | Arabia | multiple types | 10 | ? | ? | Handshape, Words, Sentences | Images, Videos, RGB | ? | Unknown |
19 | LSA-T | Argentina | translation | 103 | 14880 | 45gb | Sentence | Video(rgb) | Joint positions of active signer | Publicly Available |
20 | LSFB-CONT | Belgium | 6883 | 100 | 85000+ | ~2 TB | Word / Sentence | Videos | landmarks hand, end/start, french translation | Publicly available |
21 | LSFB-ISOL | Belgium | 400 | 85 | 50 000+ | 100Gb | Word | Videos | landmarks hand, end/start,french translation | Publicly available |
22 | WLASL | EEUU | 2000 | 119 | 21083 | ? | Word | Videos (youtube) | person bbox, dialect | Publicly Available |
Dataset information and related papers
- DGS Kinect 40 - German Sign Language (no website)
- Sign Language Recognition using Sub-Units, 2012, Cooper et al.
- Sign Language Recognition using Sequential Pattern Trees 2012, Ong et al.
- Sign Spotting using Hierarchical Sequential Patterns with Temporal Intervals 2014, Ong et al.
- RWTH-PHOENIX v1 - German Sign Language RWTH-PHOENIX v2
- Dataset paper 2012, Forster et al.
- Dataset extensions paper 2014, Forster et al
- Continuous sign language recognition: Towards large vocabulary statistical recognition systems handling multiple signers 2015, Koller et al. 4.May the force be with you: Force-aligned signwriting for automatic subunit annotation of corpora 2013, Koller et al.
- Deep Sign: Hybrid CNN-HMM for Continuous Sign Language Recognition
- TSPNet: Hierarchical Feature Learning via Temporal Semantic Pyramid for Sign Language Translation
- SIGNUM - German Sign Language
- Greek Sign Language (no website)
- Sign Language Recognition using Sub-Units, 2012, Cooper et al.
- Sign Language Recognition using Sequential Pattern Trees 2012, Ong et al.
- Sign Spotting using Hierarchical Sequential Patterns with Temporal Intervals 2014, Ong et al.
- Boston ASLLVD - American Sign Language
- Exploiting Phonological Constraints for Handshape Inference in ASL Video 2011, Thangali et al.
- A New Framework for Sign Language Recognition based on 3D Handshape Identification and Linguistic Modeling 2014 - Dilsizian - 84% accuracy
- PSL Kinect 30 - Polish Sign Language
- Polish sign language words recognition with Kinect 2013, Oszust et al.
- Some Approaches to Recognition of Sign Language Dynamic Expressions with Kinect 2014, Oszust et al.
- Recognition of Hand Gestures Observed by Depth Cameras 2015, Kapuscinski et al.
- PSL ToF 84 - Polish Sign Language
- Polish sign language words recognition with Kinect 2013, Oszust et al.
- Recognition of Hand Gestures Observed by Depth Cameras 2015, Kapuscinski et al.
- PSL 101 - Polish Sign Language (no website)
- LSA64 Argentinian Sign Language
- LSA64: an Argentinian Sign Language Dataset
- Sign Languague Recognition Without Frame-Sequencing Constraints: A Proof of Concept on the Argentinian Sign Language
- Dynamic Gesture Recognition and its Application to Sign Language 2017, Ronchetti
- SIGN LANGUAGE RECOGNITION BASED ON HAND AND BODY SKELETAL DATA 2017, Konstantinidis et al.
- Real-Time Sign Language Gesture (Word) Recognition from Video Sequences Using CNN and RNN 2018, Masood et al.
- Turkish sign language dataset
- MSR Gesture 3D - ASL Download site
- DEVISIGN G
- DEVISIGN D
- DEVISIGN L
- IIITA -ROBITA Indian Sign Language Gesture Database
- Recognizing & Interpreting Indian Sign Language Gesture for Human Robot Interaction 2010, Nandy et al.
- Recognition of Isolated Indian Sign Language gesture in Real Time 2010, Nandy et al.
- Purdue ASL Dataset
- CUNY ASL Dataset for Animation
- SignsWorld Atlas; a benchmark Arabic Sign Language database
- LSAT: Argentinian Sign Language for Translation
- LSFB-CONT: Belgian Sign Language dataset.
- LSFB-ISOL: isolated signs extracted from the LSFB-CONT dataset.
- WLASL: Word-Level American Sign Language (ASL)