Table of contents
Gleamoe Peanut 2023：Easily extract slideshows from videos, making learning and research more efficient
1. Introduction
In this era of lifelong learning, learning new knowledge is inseparable from video resources such as online courses, training and scientific research conferences. But extracting key information is often a headache. When watching academic videos such as online courses, trainings, interviews, popular science, salons, and conferences, you often encounter the problem of only videos without course slides; it is very tedious to find a certain knowledge point in the long video progress bar. I found Gleamoe Peanut 2023 , a desktop application that can extract slides from academic videos and generate PDF files, making knowledge acquisition easier and more convenient.
Software Features
- Fast and accurate: Peanut uses advanced algorithms and technologies to accurately extract the slides that appear in the video in a short period of time, improving users' learning and work efficiency.
- Multiple format support: With built-in plug-ins, Peanut supports the processing and extraction of multiple video formats, and users can freely choose the appropriate video file for format conversion and slideshow extraction.
- Ease of use: The operation of Peanut is very simple and easy to use, and users can quickly get slide extraction results without cumbersome settings and operations.
- High-quality output: The extracted slides have high definition and integrity, and can be conveniently used in various scenarios such as learning, teaching, and research.
Software advantages
-
Safe and reliable: Peanut's data processing is completed locally, avoiding the risks of data leakage and network attacks, and ensuring data security.
-
Efficient and stable: Peanut utilizes the hardware resources of the local computer to extract video slideshows faster and more stably, improving user efficiency.
-
Personalization: Peanut has rich automatic configuration and manual configuration flexibility to meet the rich personalized needs of users.
2. Software interface
The user interface is concise and clear, divided into four panels : video preview , basic information , parameter setting and output preview .
3. Installation and use
Software installation package download (Alibaba cloud disk): gleamoe-peanut-0.7.1-win.exe (https://www.aliyundrive.com/s/pGimnYeoprU)
software installation package download (AWS S3): gleamoe-peanut-0.7 .1-win.exe (https://gleamoe.s3.ap-northeast-1.amazonaws.com/gleamoe-peanut-0.7.1-win.exe)
- Installation: Click the download link above to download the software installation package, the file name of the installation package is gleamoe-peanut-0.7.1-win.exe, double-click to install. A desktop shortcut and start menu entry are automatically created at the end of the installation.
- Launch: Double-click the desktop shortcut or click the start menu item to launch.
- Exit: Click the "Close" icon in the window title bar to completely exit the software.
- Update: Gleamoe Peanut 2023 will automatically detect the update and download the upgrade package when it starts. When the download is complete, a window will pop up to remind the user. Click "Restart the program and update" to install the upgrade package file immediately; click "Process later" to start the next time Updates are installed automatically.
- Built-in plug-ins: The built-in plug-ins will be downloaded and installed automatically at the first startup, and the plug-ins can be used normally at the next startup.
For more usage details, please refer to the "Software Manual": https://gleamore.feishu.cn/docx/ZlJRd56DzoEEhlxSSAJcEkcCnJe
4. Technical details
processing logic
First determine the position of the slides in the video, define the recognition range, then compare the similarity between the slides at a certain frame interval, remove duplicate slides and merge them into PDF files. (All stages in this process allow for manual manipulation or correction.)
-
Two ways to position slides
- Auto position (default): The program can automatically detect the slide position. (Currently due to technical limitations, users need to choose whether the slide edge coincides with the video boundary.)
- Manual positioning (optional): The user can drag the video progress to any moment, click "Manual positioning" and then drag the border of the rectangular selection area to delineate the slide position.
Both approaches have pros and cons. The "automatic positioning" method is easy to operate, but sometimes it may not be able to locate accurately; the "manual positioning" method has a higher priority, and the slide positioning is accurate, but the operation is a bit cumbersome, and may introduce problems when processing the beginning and end of the video. Frames that do not contain slides need to be used in conjunction with the "start and end time" to have a better effect.
-
slide extraction
- Compare slides with similar frames, determine whether the slides overlap, and choose according to the similarity.
-
Preview operations and merging
- The extracted slides are temporarily stored locally as pictures, and users can open the picture folder through the "Operation" menu of the "Output Preview" panel to manually delete duplicate slides. "Refresh output preview" to confirm the result of manual adjustment, click "Export to PDF" to export.
-
Overwrite Mode vs Append Mode
- Overwrite mode: each time you click "Preview Output", the newly generated preview image will replace the previous preview image;
- Append mode: each time you click "Preview Output", the newly generated preview image will be appended to the previous preview image.
4. Commonly used shortcut keys
Ctrl+ O: Open the pending video
Ctrl+ S: Export the extracted slideshow
Ctrl+ D: Open the data folder
Ctrl+ Shift+ D: Empty the data folder
Ctrl+ Q: Quit the application
Space: Play/Pause
Shift+ M: Mute/Resume volume
F8: Full screen playback of the video
[: Previous frame
]: Next frame
→: Forward 1 second
Alt+ →: Forward 0.1 second
Shift+ →: Forward 10 seconds
Ctrl+ →: Forward 100 seconds
←: Backward 1 second
Alt+ ←: Backward 0.1 second
Shift+ ←: Backward 10 seconds
Ctrl+ ←: Backward 100 seconds
Shift+ C: Clear
R: Set the current time to Reference frame
I: set the current time as the start time
O: set the current time as the end time
Shift+ R: reset the time setting parameters
Ctrl+ P: open the plug-in folder
Ctrl+ R: reload
Ctrl+ 0: adjust the interface to the actual size
Ctrl+ =: zoom in
Ctrl+ -: Zoom out
Ctrl+ M: Minimize
F11: Toggle full screen
Ctrl+ Shift+ R: Force reload
Ctrl+ Shift+ I: Developer tools
Through the above shortcut keys, the user can conveniently edit and adjust the extracted slides to improve the production efficiency and effect of the slides.
5. Example of use
Take an academic report video (https://www.koushare.com/video/videodetail/29198) titled "A hydrodynamic picture of domain wall velocities" in Koushare as an example.
- First download the video (vid 29198) using KouShare-dl on GitHub :
Microsoft Windows [版本 10.0.22000.1696]
(c) Microsoft Corporation。保留所有权利。
D:\Desktop>ks login 15812345678
凭证过期，需要重新登陆。
短信验证码发送成功，请输入6位验证码：945531
登陆成功。
token文件保存成功。
D:\Desktop>ks info 29198
登陆凭证有效。
A hydrodynamic picture of domain wall velocities (vid=29198):
时长：86min 讲者：Matti Jarvinen
体积：290MB [超清] 单位：Asia Pacific Center for Theoretical Physics
日期：2022-06-28 专题：中国科学院理论物理研究所
类别：免费视频 分组：中国科学院理论物理研究所学术报告
视频简介：Physics of domain walls and bubble nucleation is important especially in view of applications to the early universe and production of gravitational waves. I demonstrate that domain walls and their movement for a class of strongly coupled theories can be described in terms of a simple four dimensional "extended" hydrodynamic model. This model puts together perfect fluid description of the strongly coupled fluid with additional scalar field which is active near the domain wall. I go on analyzing the velocity of the moving domain wall, which is usually challenging and thought to incorporate out-of-equilibrium physics. I obtain a simple formula for the velocity essentially in terms of the equation of state and perfect fluid hydrodynamics.
D:\Desktop>ks save 29198
登陆凭证有效。
A hydrodynamic picture of domain wall velocities vid=29198 超清
[>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> ] 95% 276.74MB/290.10MB 1.2MB/s
- Run Geamoe Peanut 2023, (the first time you start the software, you need to scan the QR code to log in), and open the video. In the "Parameter Settings" area, set according to the figure below, and click the "Preview Output" button.
- The processing takes about 1 minute, and the larger the video, the longer it will take. After the processing is completed, click the "OK" button in the prompt dialog box to preview the processed image.
- In the "Output Preview" area, scroll the mouse wheel to turn the page left and right; click the picture to zoom in and view the picture; right-click the picture to have several options to delete and reorder. After deleting redundant and repeated pictures and fine-tuning the order of pictures, click the "Export to PDF" button at the bottom of the "Parameter Settings" area to save the pictures in the "
Output Preview" area as a PDF file (the location of the PDF file is previously in the " Export Name" set). - View PDF files with a PDF reader and take further notes on slides, print and more.
The effect is still very good~
6. Advanced skills
- The effect is better when the resolution of the video is higher. Use high-definition or ultra-definition video whenever possible, and avoid standard-definition or low-definition video.
- If the automatic positioning of "Slide edge" is not accurate, you can try to use "Manual positioning" to circle the slide area instead of automatic positioning. At this time, if you specify the "Start and end time" at the same time, you can avoid getting more unnecessary pictures.
- "Interval Time" supports decimals, such as 2.1 seconds.
- Use Append Mode to process video in segments, step by step. Sometimes the speaker in the academic report video turns the page quickly due to time constraints and has no time to finish all the PPTs. In this case, you can use the "append mode" with the "start and end time" to process the part of the normal speech speed, and then adjust the "interval Time" to handle the quick page turning part to get a PPT preview picture that includes all pages.
More information
For more information, please pay attention to the "Gleamoe" WeChat official account!
