Ghost mannequin AI is a way to produce invisible mannequin images of clothing faster than a purely manual edit, while a trained retoucher approves every file. Alpina Production is a video production studio with AI assisted production run by humans, and we bring that split to apparel catalogs. The software handles masking, shape reconstruction and consistency. People decide whether a collar, seam or hem looks right enough to publish.
What a ghost mannequin service includes
You send photos of garments on a mannequin, on a model, or laid out flat. You receive clean images where the garment holds its shape with nothing inside it.
- Support removal: mannequin, stand, clips, pins and hands taken out of the frame.
- Interior reconstruction: the back of the neckline, sleeve openings, waistband and lining rebuilt so the garment reads as hollow and worn.
- Shape and symmetry correction: uneven shoulders, twisted seams and bunched fabric straightened without changing the cut.
- Color and light matching: one consistent look across every item in the collection.
- Wrinkle and lint cleanup: distracting creases, threads and dust removed, while real fabric texture stays.
- Exports: files sized and named for your shop, marketplaces and internal DAM.
This is an apparel photo editing service, not a replacement for the shoot. If you also need new imagery, see our work on AI product photography.
Flat lay, worn shots and the invisible mannequin effect
The invisible mannequin effect looks simple and is not. A real mannequin gives the garment a body shape, but it also hides the inside of the collar and the back panel. Flat lays are quick to shoot, yet they read as lifeless next to competitors who show volume.
Most professional workflows combine several frames: the garment on a mannequin from the front, back and sides, plus a shot of the interior. The editor then stitches the visible inner surfaces into the final image. AI tools speed up the masking and the reconstruction of hidden areas. They are weaker at details they cannot see, so a person checks anything the software had to guess.
A flat lay can also be converted, but the result depends on the garment. A structured jacket or a knit with clear form converts better than a floppy silk blouse, where the volume has to be invented. We tell you which items in a batch are safe to convert and which need a mannequin shot.
Requirements for source photos
Good source files save more time than any tool. These points matter most:
- Same setup for every item. Keep camera height, distance, lens and lighting fixed so the batch needs one look, not fifty.
- Plain, evenly lit background. Soft shadows are fine. Colored backgrounds that spill onto light fabric are not.
- Whole garment in frame with room around it. Cropped sleeves and hems cannot be rebuilt reliably.
- Sharp focus on the fabric edges, collar and seams.
- Interior shots. One extra frame of the inside neckline, back panel or lining gives the editor real data instead of a guess.
- Steamed, straight garments. Heavy wrinkles cost editing time and can distort the final shape.
- Mannequin sized for the garment, so fabric does not stretch or gape.
If you are not sure about your setup, send three or four test items before the full collection. We will tell you what to change.
Edge cases: transparency, patterns and lining
Some garments need more care, and a clear plan beats surprises late in the batch.
- Transparent and sheer fabrics. Chiffon, organza, lace and mesh must show the background or inner layers through the cloth. Automatic masking tends to turn these areas solid or leave halos. They are finished by hand against the original.
- Stripes, checks and prints. Patterns must run straight across seams and stay aligned after shape correction. We check that lines do not warp near sleeves and the neckline.
- Visible lining. If the lining shows at the neckline or inside a jacket, it has to match the real garment in color and texture. We ask for the interior shot instead of drawing it from nothing.
- Dark and glossy items. Black fabric hides seams and leather reflects the room. Both need careful light work so the garment does not turn into a flat silhouette.
- Layered and accessorized pieces. Belts, scarves, hoods and multi-layer outfits may need separate frames and a manual composite.
- Logos and labels. Text and branding are never redrawn by a generative tool. Where AI touches those areas, the editor compares the result with the source.
Deliverable specs for marketplaces
Marketplaces and fashion platforms publish their own image rules, and those rules differ and change. Instead of promising one fixed spec, we work from yours. In general terms, a delivery usually covers:
| Item | Typical approach |
|---|---|
| Background | Pure white or transparent, depending on the channel |
| Aspect ratio | One consistent ratio across the batch, often portrait |
| Resolution | High enough for zoom, with smaller web exports |
| Format | Layered PSD or TIFF master, JPEG or PNG exports |
| Color | sRGB for web, other profiles on request |
| Naming | SKU based file names and view suffixes |
| Views | Front, back and detail, set per category |
Send the current requirements of each channel in your brief. We set the export presets once and apply them to the whole collection, so a new marketplace later means one more preset, not a new edit.
How the process runs
A professional ghost mannequin production usually follows these steps:
- Brief and sample. You share the product range, the target channels and a few representative garments.
- Test batch. A handful of items covers easy and difficult cases and sets the reference look.
- Style approval. You sign off on shape, light, shadow and crop.
- Batch production. AI tools handle masking and reconstruction across the range.
- Human QC. A retoucher checks each image against the source, fixing seams, symmetry, color and detail.
- Delivery and revisions. Files arrive named and exported, and agreed revision rounds are applied.
Every image goes through a person before it reaches you. Consistency across hundreds of SKUs is the point, and it only works when someone owns the review.
What affects the price
We quote by brief, not by a list price. These factors move the cost most:
- the number of garments and views per garment;
- the quality and consistency of the source photos;
- the share of difficult items: sheer fabric, patterns, layered pieces;
- whether interior shots are provided or must be reconstructed;
- the number of export formats and channels;
- the number of revision rounds you expect.
You can see the general ways we price work on the price overview.
Who it fits
Ghost mannequin AI suits fashion brands, online retailers and marketplace sellers with a regular flow of new apparel, where a mannequin shoot is cheaper than a model shoot but the images still need to look alike across a season. It also helps teams with an archive of mixed photos who want one look before a site redesign. If your garments are highly complex or you need lifestyle imagery, we can combine this service with AI photo retouching or a full production plan.
To start, send a handful of garments and your channel requirements through the contact page. We will return a test batch and a quote based on your range.