Kling Elements is a powerful new feature that allows you to merge multiple images into a single video, making it an excellent tool for fashion-related projects. You can create a virtual model wearing real clothing or AI-generated outfits, and Kling Elements can help you achieve stunning results. Think about creating a product commercial with an AI model that looks amazingly real. You can use the video to promote your product on social media channels or for various other purposes. The idea is to get a product commercial video that costs $25/month (Kling AI subscription). 

In this tutorial, we’ll walk you through the exact steps you can take to create your first product video with a stunning AI model. 

By the end of this tutorial, you’ll learn how to:

• Curate your images
• Generate a virtual model
• Preparing clothing items
• Prepare the product images
• Structuring your prompts
• Upscale your video for commercial distribution

Let’s get right into it!

Step 1 - Curate your images

The first step in creating a fashion video is curating your images. You’ll need a model, which can be a real person or a virtual model generated using AI. Portrait images work best, especially those with a close-up of the face and a clean ackground, such as white. However, you can also use images where the face is further away or in different aspect ratios. The AI is capable of analysing the face and generating a video regardless of the image’s orientation.

If you’re using images of people of different ages, make sure to include terms like "elderly" in your prompt to get accurate results. For example, if you omit the term "elderly," the AI might generate a model that looks much younger than the reference image.

For this tutorial, we will use a paid Midjourney account to generate the image of an early 30s model with supple skin and a fresh look. Then, we will use those images to create a video of the model in Kling AI. 

Step 2 - Generate a virtual model

Go to your favourite image generator. Use Grok AI, Freepik or Midjourney. It’s best to list the features you want to have in your AI model. Here is a list of features to get you started. 

AI model features

• Tall stature
• Light brown hair
• Oriental features, fair color
• Supple and clean skin
• Slender body
• Plus-size model figure with a little bit of thickness around the waist and hips.

Go to Midjourney and click the ‘Create’ in the left navbar. 

Use a prompt similar to the one given below to generate an AI model for your product commercial. 

Prompt:

A young woman in her early 30s oriental features, big eyes, supple and clean skin, light brown hair, tall stature,  plus sized model. The photo was taken in the style of Canon EOS R5 using Kodak Portra Pro 400 film. --ar 1:1 –quality 2

Crop the image to show the face and bit of the neck only. Since we do not want Kling AI to confuse the clothing articles with the items we are about to provide to it.

‍

The results are amazingly clear. We will use this image together with clothing items and the product we intend to prompt with this AI model. 

Step 3 - Preparing clothing items

When curating clothing items, it’s best to use isolated images of the garments on a white background. This helps the AI analyse the clothing and apply it accurately to the model. For example, you can use prompts like "a lay flat image of a black bomber jacket on a white background" to generate isolated images of clothing items.

Pinterest is an excellent resource for finding inspiration and isolated images of fashion items. Simply search for terms like "Gucci jacket on a white background," and you’ll find plenty of images to use in your projects.

 However, if your product is an evening gown or any clothing item, use your product image on a white or light background. Generate the video with your product images for a showcase. 

For our example, we’ll use this Givenchy strapless dress in crepe satin for our AI model. 

However, this is not the product we are going to showcase in this AI video. It’s a clothing item that we intend for our AI model to wear while showcasing a crystal necklace. 

You can use this image to change the color of the clothing. For example, we want this dress in neon purple. Let’s prompt Midjourney to re-create this dress in a neon purple hue. 

Prompt: 

colour the dress neon purple in the attached image. White background, no model, just the dress. Shot with a Canon camera. Muted cinematic mode.

That looks good. Let’s move on and gather the images of the product we are going to showcase using the AI model we generated and the clothing item we selected. 

Step 4 - Prepare the product images

This is the important aspect of generating an AI video with the exact product that you want to showcase with the AI model. Find a crystal clear picture of your product on a white background. In our example, we are using a crystal necklace - showcasing it on the AI model standing in a ball room. 

Lastly, you can consider creating the environment for the product display in the AI video. We want to show the model Cinderella in a ballroom, standing in the middle of a crowd. We used the following prompt:

Prompt:

A lavish ballroom fit for the kings and princesses like in the story of cinderella. Muted cinematic colours. Canon RF Full-Frame Cine Lens for Red Komodo 6k. Muted cinematic mode. 

Step 5 - Structuring your prompts

Prompting is one of the most important aspects of using Kling Elements effectively. Your prompt should include four key elements: the subject (your model), the item of clothing, the action, and the environment. For example, a prompt like "a close-up of a female wearing a fashion outfit, walking on a fashion runway inside a foggy cathedral" will guide the AI to generate a video with all these elements.

If you want more control over the outfit, you can specify the exact clothing items in the prompt. For instance, "a female wearing a black leather jacket and white pants, walking on a fashion runway" will ensure the AI applies those specific garments to the model.

Let’s go to Kling AI. You’d need a basic account to create an AI video with the elements in a professional mode. 

Log in to the dashboard. Click ‘AI Videos.’

Make sure the setting is on "Image to Video," and you’ll see a new tab called "Elements." This feature allows you to upload up to four images and merge them into a single video. Under the "Hints" section, you’ll find examples of images that can be combined effectively.

When using Elements, you’ll need to type in a prompt, which is crucial for guiding the AI. The settings should be left on "Professional Mode" for the best results, even though it consumes more credits. 

You can choose a video duration of 5 to 10 seconds and select between widescreen or vertical aspect ratios. Additionally, there’s an option for a negative prompt, which tells the AI what not to include in the video, though this is optional.

Let’s tweak the settings first. Click ‘Professional mode.’ Select the 10s duration. If you want to generate more than one video, drag the slider to the right for two variations. Leave the aspect ratio to 16:9. 

Upload the photo of the model you generated, the dress image of the product and the environment in the Elements space.

Carefully structure your prompt to ensure that it generates a perfect video. Remember, describe the action, not the description of items. Think about what you want to show in the video. Use the following prompt for the example in this tutorial. 

‍

Prompt:

The model is wearing a black satin dress and a necklace, standing in a crowd of young women. The spotlight shines on her, making her necklace and her face glitter with light. Everyone is watching her. She is standing in the middle amongst the guests. Subtle motion. The camera follows her dynamically from the front. Mute cinematic mode. Shot with Canon EF 14 mm f/2.8L II USM for Red camera. 

‍

Do not hit the ‘Generate’ button just yet. As you know a prompt is important to create an image or a video. Negative prompts are way more important than regular prompts. It tells the AI engine what not to do. Use the following negative prompt for this scenario. 

Negative Prompt:

Animation, blur, shows the female model from behind, with bad lighting, and warped hand and foot motion. Blurred dresses. 

‍

Review the settings and the prompts and click the ‘Generate’ button. 

Here are the results:

Step 6 - Upscale your video for commercial distribution

The generated video needs some upscaling to make it look professional and worthy of commercial distribution, especially on social media channels. For upscaling the video, we will use vmake.com. Vmake offers a free account that allows us to upscale the video free of cost. All the other upscale services, like Magnific, do not offer trials.

Go to vmake.ai and create an account for free. Once you reach the dashboard, click ‘Video and Image Enhancer’ in the left navbar.

Click ‘Video & Image Enhancer’ in the left navbar. 

Click ‘Upload’ to load the video you created with Kling AI.

It will start upscaling the video automatically. 

Here is the video. 

That’s it, folks. Remember to experiment with the prompts, especially negative prompts. Create the next viral video displaying your product/services using the magical Kling AI. 

‍