Gallery

There are 3 categories in gallery

  1. Dataset
  2. clipper

Dataset

Creating a new dataset

  1. Login to kandula.ai.
  2. Click on the create dataset displayed in the dataset screen at the first.
  3. A blank dataset will be displayed to edit the dataset name with “Untitled” as default in the textbox.
  4. Edit the dataset name and the dataset name should be unique or an error message will be displayed “Dataset name already exists”.
  5. Choose a dataset type to start adding images, without selecting a dataset type the images cannot be added.
  6. Click on the tick located in the left top corner of the dataset beside the textbox field.
  7. When the tick is being clicked the add images button will be enabled for adding photos to the dataset.
  8. Click on add images when the button is being enabled. A modal window will be opened to add images from the collection.
  9. Choose a collection and Select images inside the collection to add images to the dataset.
  10. Click on Done and see the selected images being displayed in the datasets.

Adding of images to the existing dataset

  1. Login to kandula.ai.
  2. Choose an existing dataset to add additional images.
  3. Click on add more images from the Menu.
  4. Click on add images button a modal will be opened to add images from collection to the dataset.
  5. Choose between my collection or shared with me and select the images you want to add to the dataset.
  6. Click on done after selecting the required images.
  7. Now you will be able to see the selected images inside the existing collection.

Creating content type Video dataset

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Click on the create dataset displayed in the dataset screen at the first.
  4. A blank dataset will be displayed to edit the dataset name with “Untitled” as default in the textbox.
  5. Edit the dataset name and the dataset name should be unique or an error message will be displayed “Dataset name already exists”.
  6. Choose a dataset content type as video and any dataset type to start adding video, without selecting a dataset type the images cannot be added.
  7. Click on the tick located in the left top corner of the dataset beside the textbox field.
  8. When the tick is being clicked the add video button will be enabled for adding videos to the dataset.
  9. Add video to dataset modal will be opened.
  10. Clipper projects will be displayed.
  11. select the required clipper project, it will navigate to the clips present for the selected clipper project.
  12. Select the required clip.
  13. Click on "Done" button in the right bottom corner of the modal
  14. In few seconds you will be able to see the saved images of the clipper will be imported inside the dataset.

!!! information "Note"

  Only the saved images can be imported to the dataset from clipper. Unsaved images needs to be saved in clipper and then it can be imported to the dataset. Refer "Saving images in clipper"

Editor view for the video player

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Click on the existing dataset of video type
  4. Click on the play button placed near the swell icon.
  5. Select clip modal will be displayed.
  6. Select the required clip to navigate for the video.
  7. Click on the proceed button, when the required clip is being selected.
  8. Before clicking on the proceed button, please check if labels are being created or present in the dataset.
  9. Draw a region on the image and assign a label.
  10. Click on the start frame on the keyframe column.
  11. Select a region for tracking.
  12. Select the end frame in the keyframe column.
  13. Click on track.
  14. After few mins, when the generation of other region is being completed, the video will start automatically.
  15. When the video starts to play you will be able to see the regions from the selected start frame and end frame.

!!! information "Note"

  "Play" button will only be displayed for video content type dataset. Multiple clipper videos can be imported to one dataset.

Reset the Selected start frame or end frame

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Click on the existing dataset of video type
  4. Click on the play button placed near the swell icon.
  5. Select clip modal will be displayed.
  6. Select the required clip to navigate for the video.
  7. Click on the proceed button, when the required clip is being selected.
  8. Before clicking on the proceed button, please check if labels are being created or present in the dataset.
  9. Draw a region on the image and assign a label.
  10. When start or end frame is being selected, Next to the track button "X" will be displayed, click on the "X" the selected values will be reset.

Filtering of dataset based on types

  1. Login to kandula.ai.
  2. In the top right corner of the window, filter button will be displayed, Click on the filter button.
  3. The menu of action will be displayed in a dropdown.
  4. Select the type of dataset you want to display, the selected type will be displayed below.

Downloading of Dataset

  1. Login to kandula.ai.
  2. Choose a dataset you would like to download, click on the dataset to navigate inside.
  3. In the top right corner of the dataset, three dotted icon will be displayed.
  4. Click on the three dotted icon and see the menu of action.
  5. Click on the download as zip option, Now you will be able to see the download starting in a few seconds.

Creating a new label

  1. Login to kandula.ai.
  2. Click on a dataset and navigate inside the dataset.
  3. Click on the tags panel located on the left of the dataset.
  4. Left panel will be expanded.
  5. Enter label name to create a new label inside the create label section located above the label section.
  6. Click on the tick beside the textbox field.
  7. You will be able to see the created label in the label section.

deleting a existing label

  1. Login to kandula.ai.
  2. Click on a dataset and navigate inside the dataset.
  3. Click on the tags panel located on the left of the dataset.
  4. Left panel will be expanded.
  5. Select a label and check if the selected label is being highlighted.
  6. Click on the three dotted icon from the labels section.
  7. Menu of action will be displayed.
  8. Click on delete, The label should be removed from the label section and images inside the label will be moved to Untagged label.

!!! information "Note"

  Untagged label cannot be deleted, It’s a default label present in all datasets.

Moving images to a label

  1. Login to kandula.ai.
  2. Click on a dataset and navigate inside the dataset.
  3. Click on the tags panel located on the left of the dataset.
  4. Left panel will be expanded.
  5. Select images you would like to move the images to a label.
  6. Check if the label exists, or create the label and click on the label.
  7. The selected images will be moved to the selected label.

Search label

  1. Login to kandula.ai.
  2. Click on a dataset and navigate inside the dataset.
  3. Click on the tags panel located on the left of the dataset.
  4. Left panel will be expanded.
  5. Click on the search icon present in the label section.
  6. Search will be expanded and type text to search for the label.
  7. The searched label will be displayed below.

Invert selection

Invert selection is a selection being used for reversal of the selected labels. For e.g There are 10 labels and you would like to select 7 labels. So you can select the remaining 3 labels and click on invert selection. The selected label would be reversed. You will be able to see the unselected 7 labels being selected when invert selection is being clicked. 1. Login to kandula.ai. 2. Click on a dataset and navigate inside the dataset. 3. Click on the tags panel located on the left of the dataset. 4. Left panel will be expanded. 5. Click on the three dotted icon inside the label section. 6. The menu of actions will be displayed. 7. Click on invert selection and you will be able to see the unselected labels being selected.

Editing of photos

  1. Login to kandula.ai.
  2. Click on a dataset and navigate inside the dataset.
  3. Select an image and navigate to the image editor screen.
  4. Click on the edit tab button located beside the share button.
  5. Right pane will be opened, The labels for the dataset will be displayed in the first tab and click on the second tab for editing tools for the image.
  6. Editing tools can be used and after the use of the editing options, click on save and image would be saved to the dataset based on your edits.

Drawing of bounding box

  1. Login to kandula.ai.
  2. Click an object detection dataset and navigate inside the dataset.
  3. Select an image and navigate to the image editor screen.
  4. Click on the edit tab button located beside the share button.
  5. Right pane will be opened to display the editing tool.
  6. Click on the bounding box icon.
  7. Draw bounding box over the image and you can either create a label or select a label from the dropdown menu.
  8. After drawing the bounding box and selecting a label for the bounding box, click on save

!!! information "Note"

  Bounding box is only applicable for object detection dataset type.

Drawing of polygon

  1. Login to kandula.ai.
  2. Click an Segmentation dataset and navigate inside the dataset.
  3. Select an image and navigate to the image editor screen.
  4. Click on the edit tab button located beside the share button.
  5. Right pane will be opened to display the editing tool
  6. Click on the polygon icon
  7. Draw polygon over the image and you can either create a label or select a label from the dropdown menu.
  8. After drawing the polygon and selecting a label for the polygon, click on save

!!! information "Note"

  Polygon is only applicable for segmentation dataset type.

Selecting of the bounding box and changing label to the bounding box

  1. Login to kandula.ai.
  2. Choose a dataset and navigate inside the dataset.
  3. Select an image and navigate to the image editor screen.
  4. Click on the edit tab button located beside the share button.
  5. Right pane will be opened to display the editing tool
  6. Click on the select region icon located 1st in the toolbar.
  7. Click on the dog label shown in the image to change the label of the bounding box or you can also move the bounding box to a different points and you can also resize the bounding box.
  8. Click on save to save the latest edits being done in the image.

Moving the position of the image inside canvas

  1. Login to kandula.ai.
  2. Choose a dataset and navigate inside the dataset.
  3. Select an image and navigate to the image editor screen.
  4. Click on the edit tab button located beside the share button.
  5. Right pane will be opened to display the editing tool.
  6. Click on the drag/pan icon located 2nd in the toolbar.
  7. The image can be moved inside the canvas.

Magnifying the image (Zoom in/out)

  1. Login to kandula.ai.
  2. Choose a dataset and navigate inside the dataset.
  3. Select an image and navigate to the image editor screen.
  4. Click on the edit tab button located beside the share button.
  5. Right pane will be opened to display the editing tool.
  6. Click on the zoom in or out icon located 3rd in the toolbar.
  7. The image can be zoomed in/out using the icon or you can also zoom the image using trackpad.

Display of labels over the bounding box hide/unhide

  1. Login to kandula.ai.
  2. Choose a dataset and navigate inside the dataset.
  3. Select an image and navigate to the image editor screen.
  4. Click on the edit tab button located beside the share button.
  5. Right pane will be opened to display the editing tool.
  6. Click on the label icon located 4th in the toolbar.
  7. When the label is being selected the label will be displayed. When the label icon is being deselected the labels will not be displayed over the bounding box.

Unhiding the bounding box from the image

  1. Login to kandula.ai.
  2. Choose a dataset and navigate inside the dataset.
  3. Select an image and navigate to the image editor screen.
  4. Click on the edit tab button located beside the share button.
  5. Right pane will be opened to display the editing tool
  6. Draw the bounding box, For each bounding box drawn over the image a card would be created in the right pane.
  7. Click on the eye symbol inside the card of the bounding box needed to be hidden over the image.
  8. The box would be hidden and the eye icon will be changed to image
  9. And, if you want to see the bounding box again click on the above icon to display the bounding box.

Bounding box should be fixed

  1. Login to kandula.ai.
  2. Choose a dataset and navigate inside the dataset.
  3. Select an image and navigate to the image editor screen.
  4. Click on the edit tab button located beside the share button.
  5. Right pane will be opened to display the editing tool.
  6. Draw the bounding box, For each bounding box drawn over the image a card would be created in the right pane.
  7. Click on the lock symbol inside the card of the bounding box needed to be locked over the image. The bounding box cannot be moved.
  8. When the lock is being clicked again the bounding box can be resized and moved.

Deleting of the bounding box over the image.

  1. Login to kandula.ai.
  2. Choose a dataset and navigate inside the dataset.
  3. Select an image and navigate to the image editor screen.
  4. Click on the edit tab button located beside the share button.
  5. Right pane will be opened to display the editing tool.
  6. Draw the bounding box, For each bounding box drawn over the image a card would be created in the right pane.
  7. Click on delete button for which bounding box needs to be deleted, Since for each bounding box a separate card would be created.
  8. Click on the delete button the bounding box and the card for the bounding box would be deleted.

Denoise

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Create new dataset (Refer create dataset) or select a existing dataset.
  4. Select a image and navigate to the image editor screen.
  5. On the top right corner, there will be multiple tabs and navigate to the second tab.
  6. Click on denoise button.
  7. The image will load for a second and you will be able to see a draggable cursor in the centre of the image for comparsion of original and denoised images.
  8. Click on save. The Original image will be replaced by the denoised image.

!!! information "Note"

  The denoised image is the duplicate copy of the original image. If you require the original image, you need to add the same image from onestore folder (Refer Add images to dataset).

Super resolution

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Create new dataset (Refer create dataset) or select a existing dataset.
  4. Select a image and navigate to the image editor screen.
  5. On the top right corner, there will be multiple tabs and navigate to the second tab.
  6. Click on super resolution button.
  7. The image will load for a second and you will be able to see a draggable cursor in the centre of the image for comparsion of original and super resolution images.
  8. Click on save. The Original image will be replaced by the super resolution image.

!!! information "Note"

  The super resolution image is the duplicate copy of the original image. If you require the original image, you need to add the same image from onestore folder (Refer Add images to dataset).

Clipper

Creating a clipper project

  1. Login to kandula.ai.
  2. Click on the create clipper displayed in the Clipper screen at the first.
  3. A blank clipper project will be displayed to edit the clipper name with “Untitled” as default in the textbox.
  4. Edit the clipper project name and the clipper project name should be unique or an error message will be displayed "clipper name already exists”.
  5. When a clipper project name is given. Click on tick. Clipper project would be craeted.

Creating a new clip

  1. Login to kandula.ai.
  2. Select a existing clipper project and navigate inside a clipper project.
  3. click on the clips located in the left of clipper project.
  4. Left panel will be opened to display the source and storage path for the video.
  5. Click on the source path, click on select video to import the video from onestore for editing the video.
  6. The selected video will be displayed in the video player.
  7. Click on storage path for storing the clipped videos to the selected folder.
  8. The selected folder's name will be displayed below the storage path.

!!! information "Note"

  If the video in clip 1 has started clipping atleast one time, The source and storage path cannot be changed for clip 1.

Creating multiple clip in the same project

  1. Login to kandula.ai.
  2. Select a existing clipper project and navigate inside a clipper project.
  3. click on the clips located in the left of clipper project.
  4. Left panel will be opened to display the source and storage path for the video.
  5. Click on the "+" icon located in the clips panel.
  6. For each click, a new clip will be created in the same project.

!!! information "Note"

  Same clipper project multiple videos can be clipped. Inside a clip only one video can be clipped.

Deleting a clip

  1. Login to kandula.ai.
  2. Select a existing clipper project and navigate inside a clipper project.
  3. click on the clips located in the left of clipper project.
  4. Left panel will be opened to display the source and storage path for the video.
  5. Choose a clip to continue with deletion.
  6. Click on the three icon located in the clip panel.
  7. Menu of actions will be displayed.
  8. Click on delete clip, the selected clip will be removed from the clip section.

Clipping the video (Manual)

  1. Login to kandula.ai.
  2. Select a existing clipper project and navigate inside a clipper project.
  3. Check if the video is being added to the clip or add a video.
  4. Check if the switch below the video is in manual mode.
  5. If switch is in manual mode, marker will be opened to place the single or multiple points to trim the video.
  6. After placing the points, click on run to save the selected trimmed images in the lightbox.

Swell

  1. Login to kandula.ai.
  2. Click on a dataset and navigate inside the dataset.
  3. Click on swell wizard.
  4. A modal will be opened to show the properties for the image to be selected.
  5. Select the required properties to be applied for the image.
  6. Select onestore source path to store the output images.
  7. Select collection source path to store the output images.
  8. Click on continue.

!!! information "Note"

  Onestore and collection source path is mandatory. When onestore and collection source is not selected the continue button will not be enabled.

Collaboration

Sharing of dataset

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Select a dataset you want to share.
  4. Click on the three dotted icon displayed in the top right corner of the dataset.
  5. A menu will be displayed and click on share.
  6. A modal will be opened and enter the email address you want to share with.
  7. To decide what role people will have on your file, select Viewer or Editor.
  8. Click share.
  9. Notification will be sent to the email address entered.

!!! information "Note"

  You can let up to n number of people with view or edit permissions work on a single dataset at the same time.

Stop sharing a dataset

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Select a dataset you want to share.
  4. Click on the three dotted icon displayed in the top right corner of the dataset.
  5. A menu will be displayed and click on share.
  6. Find the person you want to stop sharing the dataset with.
  7. To the right of their name, click the delete icon.
  8. To save changes, click Save.
  9. Notification will be sent to the email address removed.

Live collaboration for owner

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Select a dataset.
  4. Select a image and navigate to the editor screen.
  5. On the right panel, navigate to the fourth tab collaborations.
  6. You will be able to see the list of all the online users. The team members should be working on the shared dataset.
  7. Click on the three dotted icon. A option will be displayed for change of editor.
  8. Click on change editor and the access will be changed to the selected user.

Live collaboration for User

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Select a dataset.
  4. Select a image and navigate to the editor screen.
  5. On the right panel, navigate to the fourth tab collaborations.
  6. You will be able to see the list of all the online users. The team members should be working on the shared dataset.
  7. Click on request button if you dont have the edit access.

!!! information "Note"

  In collaboration multiple people can view but there can be only one editor at the same time for editing the same image.

Accept of request by another user

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Select a dataset.
  4. Select a image and navigate to the editor screen.
  5. On the right panel, navigate to the fourth tab collaborations.
  6. When user send a request, you will be able to see the request in the collaboration tab.
  7. Click on accept and the edit access will be changed to the accepted user.

Reject of request by another user

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Select a dataset.
  4. Select a image and navigate to the editor screen.
  5. On the right panel, navigate to the fourth tab collaborations.
  6. When user send a request, you will be able to see the request in the collaboration tab.
  7. Click on reject and edit access will not be changed.

Dataset collaboration activity

See changes made to your dataset and keep track of who made those changes. You might see changes when someone: * Draw/edit a region * Delete a region * Change label of the region. * Super resolution or denoise.

  1. Login to kandula.ai.
  2. Click on Gallery icon and navigate to dataset from left menu
  3. Select a dataset.
  4. Select a image and navigate to the editor screen.
  5. On the right panel, To see recent changes, navigate to the third tab activity.
  6. To see older changes, scroll down on the right panel.

!!! information "Note"

  Activites will be displayed only for the selected image.