How to Create an AI Demo Video for Pet Products Using a White-Background Image

11 min read
How to Create an AI Demo Video for Pet Products Using a White-Background Image

In this tutorial, we’ll use a pet grooming brush as an example. First, we’ll create a six‑panel product storyboard in PixPix, then write video prompts using an AI agent, and finally use the Seedance 2.5 model to turn these static panels into a vertical‑format product demo video.

What does the final video showcase?

The finished video places the pet grooming brush in a natural, bright home setting and demonstrates the product through realistic scenes of cats using it.

The main shots in the video include:

  • A cat observes the pet grooming brush placed on the floor

  • The owner holds the brush and combs the cat’s fur

  • Close-up of the pet grooming brush in contact with the cat’s fur

  • Fingers operate the buttons on the back of the product

  • Close-up of pet hair collected between the brush’s teeth

  • The effect of hair being pushed out after pressing the button

These shots not only highlight the product’s appearance but also help viewers quickly grasp how to use the pet grooming brush and its hair‑removal function.

Step 1: Upload a white‑background image of the pet product

Codex 图像 2026年8月20日 16_02_45.png

Enter PixPix’s infinite canvas and upload the white‑background image of the pet product you wish to turn into a video.

For this example, we’re using a handheld pet grooming brush featuring a white-and-green color scheme. The product image clearly shows:

  • The overall shape of the brush

  • The handle’s color and structure

  • The metal bristle area

  • The buttons on the back of the product

  • The hanging hole at the base of the handle

It’s recommended to choose product images with clean backgrounds, well‑defined subjects, and clear details. The clearer the reference image, the higher the accuracy of reproducing the product’s appearance in the subsequent six‑panel storyboard.

Step 2: Enter the six‑panel storyboard instructions

After uploading the product image, enter the six‑panel storyboard instructions for the video.

Image 1 is the product reference image. Please generate a set of six‑panel pet product storyboards set in a home environment—clean, tidy, and evoking a natural, authentic lifestyle. The six panels should flow smoothly, follow a clear logical sequence, and emphasize the product’s exterior details, core functions, and specific ways to operate its buttons. You may appropriately incorporate close-ups and ultra‑close shots to clearly show the buttons, operational process, and functional effects, enabling viewers to intuitively understand the product’s features and usage. Each panel must adhere to a 9:16 vertical aspect ratio. The product must strictly replicate the reference image’s shape, proportions, and details across all frames, ensuring no distortion appears anywhere in the visuals.

Once you’ve entered the instructions, select “image mode” and wait for the AI to generate the six‑panel pet product storyboard.

6宫格.png

How should the six panels be arranged?

In this video, the six panels revolve around “product presentation—actual use—functional demonstration.”

Storyboard

Content of each panel

Key highlights presented

1

A cat sits in a home setting, with a pet grooming comb placed in front of it

Product appearance and intended users

2

The owner holds the pet grooming comb to groom the cat's fur

Actual usage of the product

3

Close-up view of the pet grooming comb near the cat’s fur

Effect of the comb teeth contacting the fur

4

Fingers press the green button on the back of the product

Button location and operating method

5

Pet hair collected by the comb teeth

Ability of the product to remove loose pet hair

6

Hair being pushed out from the comb teeth area

Demonstration of the one-touch hair‑removal function

First generate a six‑panel grid to determine the shot sequence before video production, thereby reducing issues such as changes in product form, repetitive actions, and missed features.

Step 3: Select Agent mode

After generating the six‑panel storyboard, switch to Agent mode.

Have the Agent generate corresponding video prompts based on the visual content. These prompts should describe:

  • The cat’s condition and subtle movements

  • The owner’s process of using the pet grooming comb

  • The direction and speed of the comb’s movement

  • Details of the comb teeth touching the fur

  • The action of fingers pressing the button

  • The process of loose hair being expelled from the comb teeth

  • Shot composition and camera movement

  • The product’s appearance remains consistent with the reference image.

Agent video prompt

Create a 10-second, 9:16 vertical‑format, realistic e‑commerce‑style product‑promotion video featuring a cat using a self‑cleaning grooming brush. Strictly follow the revised storyboard segments 01–06, presenting the continuous sequence: “product reveal → bristle and button interaction → smooth coat grooming → loose hair collection → proper button‑activated shedding → relaxed feline finish.”

Timeline:

0.0–1.3 seconds:

Stable medium shot with a slow push‑in. The entire grooming brush is placed diagonally on a light wooden tabletop in the foreground—this is the only full‑view product appearance throughout the clip. The bristle side faces slightly toward the camera, and the surface is completely clean, free of loose hairs. A silver tabby cat sits nearby, curiously tilting its head to look at the brush while its ears twitch gently.

1.3–2.5 seconds:

Switch to a two‑thirds side‑closeup of the brush head. An adult woman’s hand grips the sage‑green handle, while the shot simultaneously reveals the rounded metal bristles on the front and the green oval button on the back. Her thumb rests beside the button but does not press it yet. The camera pans slightly to highlight the arrangement of the bristles and the position of the button.

2.5–4.3 seconds:

The cat sits quietly on a creamy white cushion. Holding the grooming brush, she presses the front bristles flat against the cat’s shoulder and back, following the direction of hair growth, and performs two slow, smooth, gentle strokes from the upper shoulder toward the rear. The brush head remains steadily pressed against the fur—never pulling against the grain or touching the cat’s face, belly, or skin. The cat maintains a relaxed posture, showing no signs of resistance.

4.3–5.7 seconds:

The brush naturally lifts off the cat’s back, with the front bristles facing the camera. A thin layer of pale silver loose hair evenly coats the bristles; the amount of hair is realistic, avoiding exaggerated clumps. The cat remains calm in the background, and the shot subtly focuses on the details of the bristles and the collected hair.

5.7–8.0 seconds:

Strictly adhere to the mechanical motion shown in the fifth frame of the revised version. Keep the brush upright, with the handle pointing downward, and adopt a clear two‑thirds side view: the back of the brush head and the green button face rightward, while the dark gray bristle surface faces leftward.

With her right hand gripping the green handle, the woman presses the green button once with her right thumb from the back of the brush head. After the button is depressed, the cleaning pusher plate moves outward along the metal bristles, pushing the pale silver hair fragments to the tips of the front bristles. Meanwhile, her left hand approaches from the left side, carefully pinching and removing this thin layer of loose hair directly from the front bristle surface.

The removed hair must remain between the left hand and the front bristles—absolutely no hair should emerge from the back of the button, nor detach from the handle or the side of the brush head. With the button positioned on the right side and the hair collected from the left front, the motion is clear and consistent with the real mechanical structure.

8.0–10.0 seconds:

Releasing the button, the cleaning pusher returns to its original position, the metal bristles extend normally again, and the brush surface becomes clean once more. The brush is naturally placed next to the cat, with the small clump of silver hair set aside beside it as evidence of use. The cat half‑closes its eyes, and one hand gently strokes its upper back once. The camera slowly pulls back, pausing steadily for the final 0.5 seconds.

Camera transitions:

Use natural matching editing. The brush remains clean in shots 1 and 2; shot 3 marks the start of grooming; shot 4 first shows loose hair; shot 5 demonstrates the button activating and removing the hair; and shot 6 reveals the brush returning to a clean state. Camera movements primarily consist of slow pushes, subtle panning closeups, and gentle pulls back—avoid rapid cuts, rotations, bright flashes, or sudden zooms.

Sound design:

Background music features a cheerful, warm acoustic guitar, soft xylophone, and subtle hand drumming, creating a clean and adorable pet‑ad atmosphere. Simultaneously, add the gentle rustling sound of bristles gliding through fur, the crisp “click” of the button being pressed, the faint mechanical hum of the cleaning pusher moving, the slight peeling noise as hair is removed from the bristles, the soothing sound of gentle petting, and a very soft purring at the end. Product sounds should be clear, while music must never overpower the action sounds. No narration, no spoken lines, no subtitles.

Negative requirements:

Do not alter the brush’s color, shape, handle, hanging hole, bristles, buttons, or cleaning pusher mechanism; do not add extra buttons or accessories; do not allow loose hair to appear from the back of the brush head, the button, the handle, or the sides; do not place the button on the front bristle layer; do not let fingers pass through the brush head; do not show loose hair before grooming begins; do not create exaggerated giant clumps of hair; do not change the cat’s breed, pattern, size, eye color, or facial features; do not introduce a second cat; do not depict deformed paws, twisted bodies, extra hands, additional fingers, fused digits, or superfluous limbs; do not portray the cat as suffering or fearful; do not include brand names, gibberish, pricing, discounts, medical claims, hair‑removal treatments, promises of absolute painlessness, or sales figures; do not employ cartoonish, illustrative, plastic‑like, or obvious CG effects.

If you need more precise control over the final video, you can have the Agent generate separate video prompts for each of the six shots, then create corresponding scenes one by one.

Step 4: Copy the video prompt

After the Agent completes its analysis, copy the generated video prompt.

The prompt should clearly specify the main action for each shot. Avoid including too many actions in a single shot, as this may lead to issues such as unnatural hand movements, product distortion, or awkward cat behavior.

For example, you could break down the entire process into:

  1. A cat observes a pet comb.

  2. The owner picks up the pet comb.

  3. Using the comb to groom the cat’s fur.

  4. Close-up showing loose hairs collected between the comb teeth.

  5. Fingers press the hair‑removal button.

  6. Loose hairs are pushed out from between the comb teeth.

Step 5: Select the Seedance 2.5 model

Return to the six‑panel storyboard, click on the scene you wish to turn into a video, and choose the video mode.

Then complete the following steps:

  1. Paste the video prompt generated by the Agent.

  2. Select the Seedance 2.5 video model.

  3. Set the video generation duration.

  4. Set the output resolution.

  5. Adjust the aspect ratio to 9:16.

  6. Submit the task and wait for the video to be generated.

Once the video is finished, check whether the product, the cat, the hands, and the fur move naturally. If any particular shot appears noticeably off, you can regenerate just that specific scene.

How can you improve the quality of pet‑product videos?

Maintain consistent product appearance

In your prompt, explicitly require the pet comb to faithfully replicate the reference image, including:

  • White comb head housing

  • Green handle

  • Shape and arrangement of the comb teeth

  • Green button on the back

  • Overall proportions of the product

  • Hanging hole at the base of the handle

Pay particular attention to the structural relationship between the front comb teeth and the rear buttons, ensuring that AI does not alter the product design across different camera angles.

Keep the cat’s movements natural

Pets don’t need to perform dramatic actions in every shot. Subtle blinking, head turns, observing the product, or maintaining a calm sitting posture are more suitable for product demonstration videos.

During grooming, also avoid:

  • Sudden deformation of the cat’s body

  • Constant changes in fur texture

  • The comb passing through the cat’s body

  • Incorrect connection between the hand and the comb

  • The cat suddenly changing breed or color in the shot

Show button operations separately

Buttons are a key feature of this pet comb, making them ideal for close-up or extreme close-up shots when displayed alone.

In the footage, simply show fingers naturally pressing the button and floating fur being pushed out from the comb area—avoid adding complex cat movements or camera rotations at the same time.

Emphasize the before-and-after changes

The functional effectiveness of the pet comb is primarily demonstrated through changes in the fur, so it’s necessary to present the complete usage process:

Begin grooming → Comb teeth collect loose fur → Display the collected fur → Press the button → Fur is concentrated and expelled

This kind of before-and-after transformation makes it easier for viewers to understand the product’s actual function than merely showing the product rotating.

Summary

The production workflow for this set of pet product videos can be summarized as follows:

Upload a white-background image of the product → Enter six-frame storyboard instructions → Generate pet interaction scenes → Switch to Agent mode → Generate and duplicate video prompts → Select Seedance 2.5 → Set video parameters → Produce the final video

By planning the product’s appearance, pet interactions, grooming process, and button functions in advance using a six-frame layout, even with just one white-background image of the pet product, you can quickly create a vertical-format product demo video that captures an authentic user experience.

AI Image Tool Built for E-commerce Teams

For new product launches, advertising, and promotional campaigns, use AI to generate product images, scene visuals, ad creatives, and short video assets — making content production faster.