Zia Vision — An Overview
When we hear the term "customer data", we almost always think of text or numeric content — such as name, email, phone number, revenue, address, discounts and so on. Whereas, in practice, customer data could also contain a variety of images, starting from pictures of people (customers or vendors) and products to screenshots of documents and signatures. Therefore, depending on the nature of business, it may be equally important (if not more) to manage such images as part of your business data.
Consider the example of a company called Zylker Electronics, which deals with the sale of household appliances such as television, washing machine, mixers, blenders etc. When they use the CRM, they would need to upload images of products from multiple brands into their modules. This catalog would help sales people pitch the right products, model numbers and refer to suitable color options while speaking to customers. Therefore, it is crucial to get these details right in the product images, as that could influence a sales person's conversation with customers and subsequently, a customer's decision to buy the product.
So, there is a great need to upload the correct images against specific product/people records. If you upload a wrong brand, or a wrong color, for instance, it takes a lot of time and effort to go back and correct these entries— valuable time that could otherwise be spent talking to customers.
Here's where Zia Vision enters the picture. Zia Vision is a powerful data intelligence tool that is trained to review images that you upload to your CRM and validate them.
Zia can recognize and classify images and alert a user whenever an invalid image is uploaded. You can upload a range of images starting from automobiles, stationary items, sports personnels, mobile phones to houses and washing machines, to train Zia in identifying objects of various kinds. Once there is sufficient training data, you can define "desired and undesired" images so that Zia can begin validating images uploaded to the CRM. (You may use pre-trained object classifiers provided by Zoho CRM or images of your own for training purpose.)
At the end of the day, you can focus on core business activities, while Zia analyzes key images and validates them for you. In order to set up Zia Vision, there are a few simple steps that you need to follow and they are outlined below.
How Zia Image Validation Works — The Concept
As you may already know, Zia is short for Zoho's Intelligent Assistant. Zia is an AI-powered engine, and therefore requires to be "trained" to identify desired images
(or undesired images as the case may be), in order to validate them. By default, Zia is already trained in a few standard image categories such food, car, motorbikes and bicycles. Apart from this, if you wish for Zia to validate images specific to your organization, you need to "train" Zia for it. This can be done by uploading what we call "Training data". Training data involves several samples of desired (or undesired) images. For example, if you want Zia to identify a "washing machine", you will upload at least fifty to sixty images of a washing machine to train Zia about how a washing machine should look. So now, after you have "trained" Zia, if someone uploaded a TV instead of a washing machine by mistake, Zia can declare the image invalid.
This is how image validation works, and to set this up, you need to create your own
Image Validation Rule.
When you create an Image Validation Rule, you have two validation methods to choose from, based on your needs.
When to use
Complete image match
Ideal when you wish the validation to be done based on a complete match. Example,
when you want to prevent someone from uploading an image of a TV instead of a washing machine.
ideal when you wish the validation to be done based on a specific object detected in it. Example,
when you want to prevent someone from uploading an image of a front-load washing machine instead of top-load.
Now that you understand how Zia Vision works and the options available in Image Validation, let's go ahead and learn how to create an Image Validation Rule.
Setting up an Image Validation Rule
Record images that are uploaded manually or enter through CRM portals or API may sometimes be incorrect or illegible. Zia can identify these images and help the reps take action to remove or replace them.
- You can set an image validation rule for both standard and custom modules.
- You can only upload images of one object type at a time. This applies to all three (desired, undesired, pre-trained) image types.
To create an image validation rule
- Go to Setup > Zia > Vision > Get Started .
If you have created rules before, click New Image Validation Rule .
- In Create Image Validation Rule page, enter the Rule name (e.g.Validate washing machines).
Where to validate,
- Select the module from the drop-down list (e.g.: Products).
- Select the layout required. (e.g: Standard).
- Select the image field that needs to be validated.
You can either validate the record image, or a custom "image upload" field in a module.
Here, make a choice on which image field to be validated. (e.g Product Image).
- Set the criteria , either All records or Selected records (e.g.Product Category is "Washing Machine"].
In this case, the image validation rule would be applicable only for "washing machine records".
- Under Type of Validation
- Choose Match the image, if you would like to validate the image based on a complete match.
- Choose Object Detection, if you would like to validate the image based on a specific object detected in it.
Complete image match
In Image Validation based on complete image match, Zia will consider an entire image as a whole, for validation. For instance, in the case of Zylker Electronics discussed above, complete image match is ideal when you want to prevent someone from uploading an image of a TV or radio instead of a washing machine. After studying washing machines from all angles, Zia will be able to declare any other images as invalid.
If this is your requirement, under the
Type of Validation section,
- Choose Match the image.
- In Upload Training Data, make sure to upload several images of a generic washing machine.
You will be required to upload a ZIP file containing this set of training data.
Now whenever someone uploads a TV by mistake for a "Washing Machine" category record, then Zia will analyze whether the image uploaded has a "complete match" with the training data that you have uploaded. Once Zia fails to find a match, then you will see that the validation has failed and the image would not be taken into the record, but instead, sent for manual approval.
Zia Object Detection involves scanning a specific part of the image for validation. Consider the same example of Zylker Electronics. Going one step further, the washing machine department (so-to-speak) wishes to make sure someone does not add the image of a
top-load washing machine to a
front-load record by mistake. Such mix-ups could happen when images are uploaded in bulk. In this case, they could choose to validate the image based on "object detection". In any image you upload for "front-load" washing machine, Zia will look for the "front-load" portion that it is trained to detect. If it is not present, Zia could declare the image invalid. So here, the entire image is not considered for validation, but only the presence of absence of a specific "object" in it.
If this is your requirement, under the
Type of Validation section,
- Choose Detect Objects.
- In Upload Training Data, make sure to upload several images of a front-load washing machine.
You will be required to upload a ZIP file containing this set of training data.
In the example above, this Zia Vision setting is applicable for all Front load category washing machines. When a product image for a front load machine is added, Zia will check whether it contains the specific "objects" that it is trained to detect— that is the actual "front-load" portion. Only if the front-load portion (the object) is detected in an image, will Zia validate it. Otherwise, if you are uploading a top-load machine by mistake for a front load record, it will be declared invalid and will wait for your manual approval.
Note: You need the specific object to be crystal clear in every image you upload as part of the training data. If you just upload front-load machines as a whole, the training data could end up being vague. Front-load and top-load could look the same to Zia, if you upload whole images. You must upload images of the specific object in question.
Also, remember that the image validation can be done based on whether the criteria is met or not met. That is, you can also write a rule saying, "validate the image only if this object is not detected".
For instance there could be images of defective products that could escape your attention when they are uploaded in bulk. Classic examples are visible scratches or dents on a TV screen, for instance. You could upload various pictures of these scratches or dents to train Zia on what these defects look like. Then, you can write a rule that says, "validate images only when these (defective) objects are NOT detected".
This way, object detection can be used to define desired as well as undesired images.
Other scenarios where object detection is useful:
- A real estate company wants to ensure that images of "luxury villas" show a swimming pool. So you would simply attach images of swimming pools as the training data for "object detection". So, Luxury Villa images will be validated only if these swimming pool objects are detected. If they are not detected, Zia will declare the image as invalid.
- A car dealership company wants to keep images of defective cars out of their records. So, closeup pictures of dents and scratches could be added as the training data for object detection. Images will be validated only if these objects are NOT detected in the whole image. If they are detected, the image will be declared invalid by Zia.
- A window cleaning/repair company wants their cleaners to upload pictures of perfect windows after a repair job on a custom image field called "Final Look". To get this done, they could upload close-up images of broken/dirty windows to train Zia on what is NOT desirable. So only if such objects are not detected, will the image be valid. If these scratches/broken glass objects are detected, the image will not be validated by Zia.
Once you are done with selection of validation type and uploaded corresponding training data, the next step is to note the action on validation failure.
- Under Action on Failure, you have a single action.
Whenever an image fails validation, it will be sent for manual approval.
The associated record will be created, however the image will wait under the "My Jobs" module to be approved, rejected or updated manually by your sales teams.
Testing the model
With the Test the model feature, the admin can assess the pattern Zia has created based on the training data . It can instantly provide results whether the give input matches with the model Zia has created and show if it is a success or failure. Based on these results the admin can make changes to the model.
To test a model,
- Go to Setup > Zia > Vision .
- Select the rule for which you want to test the model.
Test the model.
- Upload the image that needs to be validated by Zia.
- Zia will validate the image and deliver the result.
Zia will update what is called the "Success rate", periodically, based on how accurate Zia's image validation has been. If an undesired image is wrongly accepted by Zia as a suitable image, then the user has a way to convey to the system that the image validation is wrong. Similarly, if an image allowed to stay in the system, Zia understands this to be a successful validation. As this pattern repeats, Zia will maintain a success score, which tells you how effective AI image validation has been for your CRM usage. If there is a poor success rate, you may want to revisit your training data set, check the Accuracy score of your training data and realign Zia to understand desired and undesired images.
Record and Image approvals from My Jobs
The records or images can be approved from the
tab. The admin must approve the record only then can the image be approved. If the record is rejected the image is automatically disqualified.
Delete or deactivate a rule
You can delete or deactivate a rule if required. If you chose to approve the records that enter the rule, then upon deleting the rule you will be notified about the records that are awaiting an approval. Similarly, upon deactivating a rule the existing records that are awaiting an approval will be retained as is, however new records will not be allowed to enter the rule.
Enabling permission to configure image validation rule and review images
CRM administrators who have the permission to manage configuration or review images enabled in their profile will be able to create image validation rule and review images respectively. You can give the manage configuration permission to selected users who will only create the rule and not be responsible for reviewing the images and similarly the reviewers can be restricted from viewing or creating the configuration.
Under Profiles > Zia > Image Validation , there are two permissions which allow the user to do the following:
- Manage configuration: Users with this permission can only create, edit, view, enable, disable or delete a rule. They cannot review the images.
- Manage action: Users with this permission can review the images and approve or reject them. These users will not be allowed to view or create a rule.
Remember that when you toggle on Image validation for a profile, manage action will be automatically enabled. However, you have to select manage configuration manually.
To give profile permission
- Go to Setup > Users and Control > Security Control and choose a profile.
Only admin profiles or its equivalent clone can be selected.
- In Profile , go to Setup Permissions and click Zia .
- Under Zia , toggle on Image Validation .
Manage action will be enabled automatically. Click to disable.
- Select Manage Configuration or/and Manage Action .
Best practices for uploading images
Take a look at the
below that highlight the possible problems and the ways to troubleshoot them:
Case 1. A stationary supplier wants to validate images of pens and pencils received through bulk orders from the vendors. They upload the images of these objects as classifiers. However, when bulk orders are received most images of biros (ballpoint pens) are marked invalid.
The training data consists of images of all kinds of pens such as fountain pens, gel pens, marker pens, and pencils except ballpoint pens. Therefore, though Zia can identify a variety of pens it cannot identify a ballpoint pen and shows it as an invalid image. Hence, in order to get accurate result it is important to upload all images of all the objects that you want to be classified.
A mobile phone dealer wants to validate images of smart phones. He uploads pictures of all the brands of smartphones he sells, but still some images of mobile phones are shown as invalid.
The dealer uploads front images of all brands of smartphones. However, some images that are taken from other angles are difficult to identify and draw similarity between objects. Therefore, it is essential to upload images of objects from various angles for ease of identification and accuracy of results.
A distributor of toys and puzzles is confused because images of geometric or shape puzzles are identified as invalid objects though the same type of images have been used as classifiers.
A geometric shape appears similar from every angle thus uploading image of an object from either angles would always give accurate results. However, there must be enough images to draw similarity between objects. Hence, while uploading images of objects that may appear similar from all angles it is necessary to have at least 5 to 6 images (minimum requirement for Zia's image validation) of the object. We recommend to repeat one image 5 times for the system to draw correlation.
Points to remember
- If you want to upload images from your desktop, you must follow these guidelines to achieve the best results from Zia:
- The images must be in these format: JPG, JPEG, PNG, GIF, BMP, and TIFF.
- The training data should be nearly similar to the data that needs to be validated. That is, images of villas, motorbikes, cars etc. should be clear and easily identifiable for accurate validation.
- In general, the training data should have images from multiple angles, resolutions, and backgrounds for variety.
- Vision models generally cannot recognize patterns that humans cannot. So if a human cannot recognize a pattern by looking at the image for around half to one second, the model probably cannot be trained to do it either.
- You can upload as many images as you want for a better accuracy, but minimum 5 images of a particular category must be uploaded for Zia to validate images.
- Do not combine images from different categories. Use images that best depict your category.
- The training accuracy score will be impacted if these guidelines are not followed.
- We recommend you to select record or image approval as the action until you are sure of the results shown by Zia.
- We calculate the accuracy of the image validation rule by evaluating the points outlined in the guidelines. Only if a rule has a training accuracy above 80% the system will allow you apply it.
- In case multiple rules are configured, a record will enter a rule based on the configuration that matches.