i need databases for hand gesture recognition

I am a master student, and I need someone whose work just like mine to share information [ hand gesture recognition based on high density surface electromyography signals]

Answer 1

[Image Analyst]

Go here to find groups working on gesture recognition:

http://www.visionbib.com/bibliography/contentspeople.html#Face%20Recognition,%20Detection,%20Tracking,%20Gesture%20Recognition,%20Fingerprints,%20Biometrics

Find one you like and contact them to see if they have any videos to share or MATLAB code to share.

[Alyaa Ghazi]

Thank you

Answer 2

[Drew]

This answer focuses on the question header "i need databases for hand gesture recognition", that is, on searching for databases.

Update:

As pointed out in the comments, the generative AI quoted below makes quite a lot of mistakes, consistent with the caveats at the start of the answer. Even when the dataset links provided by the AI are wrong, one can do a google search with some of the keywords from the AI and find the correct link for some of the datasets.

For searching for databases, another source is https://datasetsearch.research.google.com/. A search for "Hand Gesture Recognition" came back with 100+ databases. A search for "Hand Gesture Recognition based on high density surface electromyography signals" came back with one database "Open Access Dataset and Toolbox of High-Density Surface Electromyogram Recordings" https://physionet.org/content/hd-semg/1.0.0/ . Adjusting the search terms should narrow the search in other ways.

Original Answer:

This is a good question to ask a large language model. The information below is from a large language model, so it may or may not be correct. It needs to be checked, but it may provide some good leads. .

Question: What are some publicly available databases for research on hand gesture recognition?

AI: Hand gesture recognition is a popular area of research in computer vision and human-computer interaction. Several public databases have been created to facilitate research in this field. Here are some of the most commonly used datasets:

1. 20BN-jester Dataset

• Description: This dataset is specifically designed for hand gesture recognition in the context of human-computer interaction. It contains 148,092 labeled video clips that show humans performing pre-defined hand gestures.
• Link: 20BN-jester Dataset

2. NVGesture Dataset

• Description: Created by NVIDIA, this dataset is intended for studying hand gesture recognition using multimodal data (RGB, depth, and IR). It contains 1,532 video clips performed by 20 subjects under various lighting conditions.
• Link: NVGesture

3. MSRA Hand Gesture Database

• Description: Provided by Microsoft Research Asia, this dataset contains 76,375 frames collected from 9 subjects performing 12 dynamic hand gestures.
• Link: Not directly available for download. You may need to contact the authors or Microsoft Research to gain access.

4. Cambridge Hand Gesture Dataset

• Description: This dataset includes a series of hand gesture image sequences that contain both 2D and 3D hand tracking information, suitable for gesture recognition tasks.
• Link: Cambridge Hand Gesture Dataset

5. DVS128 Gesture Dataset

• Description: This is a dataset of hand gestures recorded using a Dynamic Vision Sensor (DVS), an event-based camera that captures pixel changes in brightness instead of traditional frames.
• Link: DVS128 Gesture Dataset

6. ChaLearn LAP IsoGD Dataset

• Description: This dataset was created for the ChaLearn Large-scale Isolated Gesture Recognition Challenge and includes 47,933 videos of 21 different hand gestures.
• Link: ChaLearn LAP IsoGD Dataset

7. EgoGesture Dataset

• Description: This dataset is collected for egocentric (first-person) vision-based hand gesture recognition and contains 2,081 RGB-D video sequences and 24,161 gesture instances.
• Link: EgoGesture Dataset

8. RWTH-PHOENIX-Weather 2014T

• Description: While primarily a dataset for sign language recognition, it contains continuous hand gestures and can be useful for hand gesture recognition tasks.
• Link: RWTH-PHOENIX-Weather 2014T

When using these datasets, it's important to check the terms of use and cite the appropriate papers if you're using them for research purposes. Additionally, some datasets may require you to request access or agree to certain conditions before you can download them.

Question: Do you know of any hand gesture recognition based on high density surface electromyography signals?

AI: High-density surface electromyography (HD-sEMG) is a technique used to record the electrical activity produced by skeletal muscles. It provides detailed information about muscle activity and can be used for hand gesture recognition by interpreting the patterns of muscle activation.

There are datasets specifically collected using HD-sEMG for the purpose of hand gesture recognition. Here are a few examples:

1. Ninapro Database

• Description: The Non-Invasive Adaptive Prosthetics (Ninapro) database is a resource for studying the myoelectric activity of muscles during different hand movements. It includes HD-sEMG recordings for various hand gestures.
• Link: Ninapro Database

2. CapgMyo Dataset

• Description: The CapgMyo dataset has been created by the Chinese Academy of Sciences and includes HD-sEMG data for hand gesture recognition. It contains several datasets with different numbers of gestures and subjects.
• Link: CapgMyo Dataset

3. BioPatRec Dataset

• Description: This is a publicly available dataset associated with the BioPatRec project, which is an open-source platform for pattern recognition of myoelectric signals. The dataset includes sEMG signals for different limb positions and movements.
• Link: BioPatRec

4. MYO Armband Dataset

• Description: While not high-density in the traditional sense, the MYO armband captures EMG signals from the forearm and is used for gesture recognition. Researchers have used it to collect datasets for hand gesture recognition.
• Link: Dataset availability may vary, as it is often collected by individual research groups. You may find relevant datasets by searching for academic publications that used the MYO armband.

When working with HD-sEMG datasets for hand gesture recognition, it's essential to have a good understanding of signal processing and pattern recognition techniques, as the raw data can be quite complex. Researchers typically use machine learning algorithms to classify the different gestures based on the sEMG signal patterns.

If this answer helps you, please remember to accept the answer.

[Image Analyst]

Looks like only 3 of those links are valid.

[Walter Roberson]

And one of them is a link to a completely different dataset.

[Alyaa Ghazi]

thank u