Fitness Exercises Dataset

This dataset contains 1300+ samples of exercises with body parts, target muscles, secondary muscles and instructions. Each exercise has also a GIF url.

For the GIFs not loading:

I bought them for $1 each so that I can give them to you with 75% off! The FULL 1324 GIFs package for only the first 5 to grab them (ONLY 1 LEFT!) at that price from https://omarxadel.gumroad.com/l/exercisesdb

https://public-files.gumroad.com/7hxdwndfkmpyxb5r9k7g0yh07iqz" alt="">

✔ High Quality GIFs

✔ The whole thing (1324 GIFs)

✔ Sorted as they are in the database so you'll only need to run 1 job to change the names

✔ No watermarks

And they're limited. Make sure to get them before anyone else.😄

This dataset was created by myself. This dataset contains videos of people doing workouts. The name of the existing workout corresponds to the name of the folder listed.

Video format: .mp4 Some of the videos are muted

What is the videos resolution? The resolution of this video varies greatly, but I'm trying to find the best possible resolution so that you can lower the resolution according to what you will use later.

How about the duration of the videos? It also varies, but there is at least 1 rep on each video

What are the data sources? Mostly sourced from YouTube, but I also create some of it by myself with my friends

Need the extracted frame of each video? Try check my other dataset for the images of workout/exercise here

I needed a dataset of gym exercises, the muscles targeted by them, the equipment used and a brief description of each exercise for my project- however, I was unable to find a dataset like this anywhere- so I created one with data pulled from bodybuilding.com .

This dataset contains 470 gym exercises, links providing a description, the muscles targeted by them, the equipment used and a brief explanation of each equipment. Think of it as an all-you-need dataset either for any gym exercise related projects or for creating your workout program.

Happy Kaggling!

This dataset is a comprehensive list of gym exercises that can be used to improve your fitness. It includes exercises for all levels of fitness, from beginners to advanced. The dataset also includes information on the muscles worked by each exercise, the equipment needed, and how to do the exercise safely.

This dataset can be used to create a personalized workout routine that meets your individual fitness goals. You can use the information in the dataset to choose exercises that target the muscles you want to strengthen or tone. You can also use the information to find exercises that are safe for your fitness level.

The dataset is a valuable resource for anyone who wants to improve their fitness. It can be used by beginners to learn the basics of gym exercises, by intermediate exercisers to find new and challenging exercises, and by advanced exercisers to fine-tune their workouts.

Here are some additional tips for using the dataset:

Start with a few exercises and gradually add more as you get stronger. Listen to your body and don't push yourself too hard. Warm up before you start your workout and cool down afterwards. Stay hydrated by drinking plenty of water. Eat a healthy diet to support your fitness goals.

https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F7974466%2F69e65ffa17d835dc093fc520999e492c%2FLeonardo_Kino_XL_Create_a_vibrant_and_dynamic_digital_illustra_2.jpg?generation=1728979114108033&alt=media" alt="">

This dataset provides detailed information on 50 diverse exercises designed to promote overall health and fitness. It includes a wide range of activities suitable for beginners to advanced fitness enthusiasts, targeting various muscle groups and fitness goals. The data can be used for personal fitness planning, workout app development, or data analysis projects in health and sports science.

Column Descriptions

1. Name of Exercise: The common name of the exercise. Type: String Description: Unique identifier for each exercise in the dataset.

2. Sets: The recommended number of sets for the exercise. Type: Integer Description: Indicates how many times the group of repetitions should be performed.

3. Reps: The recommended number of repetitions per set. Type: Integer Description: Specifies how many times the exercise should be performed in each set.

4. Benefit: The primary health or fitness benefit of the exercise. Type: String Description: Briefly explains the main advantage or target of the exercise.

5. Burns Calories (per 30 min): Estimated calorie burn for a 30-minute session. Type: Integer Description: Approximates the number of calories burned by an average person (155 lbs/70 kg) performing the exercise for 30 minutes.

6. Target Muscle Group: The main muscles or muscle groups engaged during the exercise. Type: String Description: Lists the primary muscles worked, helping users target specific areas.

7. Equipment Needed: Any equipment required to perform the exercise. Type: String Description: Specifies necessary equipment, or "None" if the exercise can be performed without equipment.

8. Difficulty Level: The relative challenge level of the exercise. Type: String Description: Categorizes exercises as "Beginner," "Intermediate," or "Advanced" to guide appropriate selection based on fitness level.

This fitness dataset provides personalized exercise plans tailored to individuals' physical features, activity levels, and gender. It is designed to assist individuals in achieving their fitness goals by offering customized workout routines that optimize effectiveness and safety.

Key Features:

Physical Features: The dataset includes physical attributes such as height (h), weight (w), body mass index (BMI), body fat percentage, muscle mass, and other relevant metrics. These features are crucial for determining an individual's baseline fitness level and guiding exercise recommendations. Gender: Gender is an essential factor in designing personalized exercise plans. The dataset categorizes individuals into different gender groups to account for physiological differences and tailor workouts accordingly. Activity Levels: The dataset captures information about individuals' activity levels, including their daily physical activity, exercise frequency, intensity, and duration. Understanding activity levels helps in prescribing appropriate workout regimens that align with individuals' lifestyles and fitness goals. Exercise Preferences: Individuals may have preferences for specific types of exercises, such as cardio, strength training, flexibility, or endurance activities. The dataset includes information about exercise preferences to ensure that recommended workout plans are enjoyable and sustainable. Fitness Goals: The dataset allows individuals to set personalized fitness goals, such as weight loss, muscle gain, improved endurance, or overall health and wellness. Exercise plans are tailored to help individuals achieve their specific objectives effectively and efficiently.

It sounds like you have a substantial amount of personal exercise and health data accumulated over 150 days. This data can provide valuable insights into your fitness journey and overall well-being. Here are some suggestions on how you can analyze and make the most of this information:

Exercise Types:

Identify the types of exercises you've been engaging in. Categorize them into cardiovascular, strength training, flexibility, and other categories. Note the frequency and duration of each type of exercise.

Intensity Levels: Assess the intensity of your workouts. This can be measured in terms of heart rate, perceived exertion, or weight lifted. Determine if there are patterns in intensity levels over time.

Progress and Setbacks: Look for trends in your progress. Are you consistently improving, or have you encountered any setbacks? Identify factors that contribute to your success or challenges.

Rest and Recovery: Analyze your rest days and recovery strategies. Ensure that you're allowing your body enough time to recover between intense workouts. Look for patterns in your energy levels and performance related to rest.

Nutrition and Hydration: Correlate your exercise data with your nutrition and hydration habits. Consider whether certain eating patterns impact your workouts positively or negatively.

Sleep Patterns: Examine your sleep data if available. Adequate sleep is crucial for recovery and overall health. Identify any correlations between your sleep patterns and exercise performance.

Mood and Stress Levels: Reflect on your mood and stress levels on different days. Exercise can have a significant impact on mental well-being. Consider whether there are connections between your exercise routine and your emotional state.

Injury Analysis: If you've experienced any injuries during this period, analyze the circumstances surrounding them. This can help in understanding potential risk factors.

Goal Alignment: Evaluate whether your exercise routine aligns with your initial goals. Are you progressing toward your desired outcomes?

Adjustment of Exercise Routine: Based on the analysis, consider adjustments to your exercise routine. This might involve modifying the types of exercises, intensity, or frequency.

Remember, the goal of analyzing this data is to make informed decisions about your fitness routine, identify areas of improvement, and celebrate your successes. If you have specific questions about the data or need guidance on certain aspects, feel free to provide more details for personalized advice.

🏋️ About This Dataset This comprehensive fitness dataset contains over 600,000 structured workout routines and exercise entries scraped from fitness planning platform data. The dataset includes both detailed exercise-level data and program-level summaries, making it ideal for building recommendation systems, analyzing workout patterns, and understanding fitness program structures.

📊 Dataset Overview Two complementary files: 1. Main Dataset (fitness_exercises.csv): 605,033 individual exercise entries with detailed workout information 2. Program Summary (program_summary.csv): 2,598 unique fitness programs with aggregated metadata

🔑 Key Features

Main Dataset (605K+ rows): - Exercise details: name, sets, reps, intensity - Program structure: week, day, time per workout - User targeting: fitness level, goals, equipment needs - Temporal data: creation and edit timestamps - Program metadata: length, number of exercises per workout

Program Summary (2.6K+ programs): - Program overview: title, description, fitness level - Target goals and equipment requirements - Program duration and workout timing - Total exercise count per program - Creation and modification timestamps

🎯 Use Cases - Building workout recommendation systems - Analyzing fitness program effectiveness and popularity - Understanding exercise patterns and program structures - Creating personalized workout generators - Fitness app development and research - Program-level analysis and clustering

🔧 Technical Details - Format: CSV files - Combined size: ~300MB+ - Data quality: Minimal missing values (<1% for most columns) - Collection period: [Add your scraping date] - Source: Fitness platform data (with attribution)

📝 Data Dictionary

Main Dataset: - title: Workout/program name - description: Detailed workout description - level: Fitness level (beginner/intermediate/advanced) - goal: Primary fitness objective - equipment: Required equipment type - program_length: Duration in weeks - time_per_workout: Duration per session (minutes) - week/day: Position in program structure - exercise_name: Specific exercise name - sets/reps: Exercise volume parameters (negative values are time in seconds) - intensity: Exercise intensity level

Program Summary: - title: Program name - description: Program overview and objectives - level: Target fitness level - goal: Primary fitness goal - equipment: Required equipment - program_length: Total program duration (weeks) - time_per_workout: Average workout duration (minutes) - total_exercises: Total number of exercises in program - created/last_edit: Program timestamps

🔗 Data Relationship The program_summary file provides aggregated views of the detailed exercise data, allowing for both micro-level exercise analysis and macro-level program insights.

⚖️ Important Notes - Data collected from publicly available fitness planning platform - Cleaned and structured for research/educational use - Please respect original platform's terms of service - Consider this for non-commercial research and educational purposes

Content

This dataset represents sequential poses that can be used to distinguish 5 physical exercises. The exercises are Push-up, Pull-up, Sit-up, Jumping Jack and Squat. The dataset consists of 33 landmarks that represents several important body parts' positions. Using these landmarks, the angles and the distances between several landmarks are calculated and included in the dataset. The sequence of the poses is provided by preserving the frame order in every record.

https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F5365924%2Fb8c7ec50ebc270533628c7d05966ecbd%2FScreenshot%202023-02-22%20at%2020.30.37.png?generation=1677087060116097&alt=media" alt="">

Collection Process

About 500 videos of people doing the exercises have been used in order to collect this data. The videos are from Countix Dataset that contain the YouTube links of several human activity videos. Using a simple Python script, the videos of 5 different physical exercises are downloaded. All the frames of the videos are extracted, processed and included in the dataset.

Processing Data

For every frame, MediaPipe framework is used for applying pose estimation, which detects the human skeleton of the person in the frame. The landmark model in MediaPipe Pose predicts the location of 33 pose landmarks (see figure below). Visit Mediapipe Pose Classification page for more details. Using these landmarks, the angles and the distances between several landmarks are calculated.

https://mediapipe.dev/images/mobile/pose_tracking_full_body_landmarks.png" alt="33 pose landmarks">

This Gym Exercise Dataset offers a comprehensive examination of various exercises and their detailed components. It focuses specifically on exercises performed using machines commonly available in gym settings.

The dataset encompasses: - Detailed breakdowns of machine-based exercises - Specific components and parameters for each exercise - Information on proper form and technique - Data on muscle groups targeted by each exercise

This collection serves as a valuable resource for: - Fitness professionals developing evidence-based training programs - Researchers studying exercise biomechanics and efficiency - Gym equipment manufacturers interested in user interaction data - Data scientists exploring patterns in exercise routines and preferences

A gym workout/exercise video dataset for video classification problems. This dataset was created and processed by our team and a part of the raw data from Workout/Exercises Video of Hasyim Abdillah. This dataset contains videos of people doing workouts, the name of the existing workout corresponds to the name of the folder listed. All videos are in .mp4 format.

There are 3 parts in the dataset: - raw_data: - data-btc: include 652 videos from [Workout/Exercises Video] - data-crawl: 334 videos crawled from YouTube. - verified_data: include the preprocessed data. We have removed noise from the raw data and divided it into smaller videos (maximum 10 - 13s for a video). - data_btc_10s: 817 processed videos from Workout/Exercises Video of Hasyim Abdillah. - data_crawl_10s: 754 processed videos from our crawl data. - test: include 61 videos for testing (video will have noises for better-evaluating model).

You can visit my Github Repo here for a complete guide to using VideoMAE (a video classification model) with this data.

This dataset contains a collection of GIFs demonstrating the lunge exercise from different perspectives (front and side views). Each sample shows pose analysis and angle visualization, making it useful for computer vision tasks such as:

• Exercise form detection
• Human pose estimation
• Workout analysis

The GIFs were collected from publicly available sports videos on YouTube and then converted into GIF format for easier use in computer vision pipelines.

This comprehensive exercise pose dataset consists of 2701 rows and 133 columns, providing valuable insights into human skeleton movements during various exercises. With a focus on analyzing different exercise poses, the dataset encompasses seven distinct classes: Rest, Left Bicep, Right Bicep, Left Shoulder, Right Shoulder, Left Tricep, and Right Tricep.

Each row corresponds to a specific exercise, while the columns represent different aspects of the human skeleton model. The dataset captures the coordinates (X, Y, Z), and visibility values of 33 landmarks, resulting in a total of 132 values per exercise. These landmarks serve as key reference points to evaluate body positions and movements accurately.

The dataset is invaluable for researchers, data scientists, and fitness enthusiasts seeking to understand human skeleton kinetics during exercise routines. It enables comprehensive analysis of body posture, movement patterns, and joint angles, facilitating in-depth insights into exercise performance and form.

This dataset is an ideal resource for machine learning enthusiasts who wish to develop and evaluate models for exercise pose recognition, gesture analysis, and exercise tracking systems. The rich annotation and class labels make it suitable for tasks such as activity recognition, pose estimation, and exercise recommendation systems.

By contributing this dataset to Kaggle, I aim to foster collaboration, knowledge sharing, and further advancements in the analysis of human skeleton movements. Researchers and data scientists can leverage this dataset to devise innovative approaches and algorithms to improve fitness monitoring, performance tracking, and personalized exercise guidance.

Dataset Name: MEx Multi-modal Exercise dataset

Description: The MEx Multi-modal Exercise dataset contains data of 7 different physiotherapy exercises, performed by 30 subjects recorded with 2 accelerometers, a pressure mat, and a depth camera.

Application: The dataset can be used for exercise recognition, exercise quality assessment, and exercise counting, by developing algorithms for pre-processing, feature extraction, multi-modal sensor fusion, segmentation, and classification.

Data Collection Method: Each subject was given a sheet of 7 exercises with instructions to perform the exercise at the beginning of the session. At the beginning of each exercise, the researcher demonstrated the exercise to the subject, then the subject performed the exercise for a maximum of 60 seconds while being recorded with four sensors. During the recording, the researcher did not give any advice or keep count or time to enforce a rhythm.

Sensors: 1. Obbrec Astra Depth Camera - Sampling frequency: 15Hz - Frame size: 240x320

1. Sensing Tex Pressure Mat

   • Sampling frequency: 15Hz
   • Frame size: 32x16

2. Axivity AX3 3-Axis Logging Accelerometer

   • Sampling frequency: 100Hz
   • Range: 8g

Sensor Placement: All exercises were performed lying down on the mat while the subject wore two accelerometers on the wrist and the thigh. The depth camera was placed above the subject facing downwards, recording an aerial view. The top of the depth camera frame was aligned with the top of the pressure mat frame and the subject’s shoulders, ensuring the face would not be included in the depth camera video.

Data Folder: The MEx folder contains four folders, one for each sensor. Inside each sensor folder, 30 folders can be found, one for each subject. In each subject folder, 8 files can be found for each exercise with 2 files for exercise 4 as it is performed on two sides. (Subject 22 will only have 7 files as they performed exercise 4 on only one side.) Each line in the data files corresponds to one timestamped sensory data.

Attribute Information:

• Accelerometer (act and acw files):

  1. Timestamp
  2. X value
  3. Y value
  4. Z value
  5. Min value: -8
  6. Max value: +8

• Pressure Mat (pm file):

  1. Timestamp 2-513: Pressure mat data frame (32x16)
  2. Min value: 0
  3. Max value: 1

• Depth Camera (dc file):

  1. Timestamp 2-193: Depth camera data frame (12x16)
  2. Depth camera data frame is scaled down from 240x320 to 12x16 using the OpenCV resize algorithm
  3. Min value: 0
  4. Max value: 1

Note:

Because this dataset has been used in a competition, we had to hide some of the data to prepare the test dataset for the competition. Thus, in the previous version of the dataset, only train.csv file is existed.

Content

This dataset represents 10 different physical poses that can be used to distinguish 5 exercises. The exercises are Push-up, Pull-up, Sit-up, Jumping Jack and Squat. For every exercise, 2 different classes have been used to represent the terminal positions of that exercise (e.g., “up” and “down” positions for push-ups).

Collection Process

About 500 videos of people doing the exercises have been used in order to collect this data. The videos are from Countix Dataset that contain the YouTube links of several human activity videos. Using a simple Python script, the videos of 5 different physical exercises are downloaded. From every video, at least 2 frames are manually extracted. The extracted frames represent the terminal positions of the exercise.

Processing Data

For every frame, MediaPipe framework is used for applying pose estimation, which detects the human skeleton of the person in the frame. The landmark model in MediaPipe Pose predicts the location of 33 pose landmarks (see figure below). Visit Mediapipe Pose Classification page for more details.

https://mediapipe.dev/images/mobile/pose_tracking_full_body_landmarks.png" alt="33 pose landmarks">

Fitness exercises

https://www.melarossa.it/wp-content/uploads/2019/11/30-days-superhero-fitness-challenge-750x375.jpg?x75642" alt="aa">

This dataset contains all the information related to fitness exercises. The exercises are split in the following categories:

• 'back',
• 'cardio',
• 'chest',
• 'lower arms',
• 'lower legs',
• 'neck',
• 'shoulders',
• 'upper arms',
• 'upper legs',
• 'waist']

Dataset

This dataset was created by Talha Anjum

Released under Apache 2.0

Contents

This dataset contains detailed employee engagement survey responses collected voluntarily from employees of Pierce County Government in Washington State. The survey measures employees’ agreement levels on various workplace statements to assess overall engagement and satisfaction.

Key Features

• Survey Responses: Numeric values ranging from 0 to 4, where 0 means Not Applicable (N/A), 1 = Strongly Disagree, 2 = Disagree, 3 = Agree, and 4 = Strongly Agree. Each response reflects the employee’s sentiment toward a specific engagement question.
• Demographics & Hierarchy: Includes information on the employee’s role, department, and organizational structure with identifiers for director, manager, supervisor, lead, and staff.
• Survey Questions: Detailed statements about workplace environment and employee engagement, allowing deep analysis of opinions.
• Time Frame: Survey data covers responses collected in specific years (column Year).

Recommended Analyses

• Identify which questions employees most strongly agree or disagree with.
• Explore trends and differences in responses by department, role, or managerial hierarchy.
• Provide insights for management to improve employee satisfaction and engagement.

Source

The dataset was provided by Pierce County, WA. Licensed under Public Domain.

Dataset

This dataset was created by Ziya

Released under CC0: Public Domain

Contents

Overview

https://i.imgur.com/ZUX61cD.png" alt="Overview">

Context

The method of disuniting similar data is called clustering. you can create dummy data for classifying clusters by method from sklearn package but it needs to put your effort into job.

For users who making hard test cases for example of clustering, I think this dataset helps them.

Try out to select a meaningful number of clusters, and dividing the data into clusters. Here are exercises for you.

Dataset

All csv files contain a lots of x, y and color, and you can see above figures.

If you want to use position as type of integer, scale it and round off to integer as like x = round(x * 100).

Furthermore, here is GUI Tool to generate 2D points for clustering. you can make your dataset with this tool. https://www.joonas.io/cluster-paint

Stay tuned for further updates! also if any idea, you can comment me.