One furniture photo, three room settings

A DTC furniture brand generates living room, bedroom, and home study lifestyle images from a single studio product shot without booking a second shoot. Cnaps studio segments the furniture automatically, handles the shadow and background compositing, and produces three room scene variants in one pipeline run.

Why do furniture brands need lifestyle imagery beyond the studio shot?

A furniture product shot on a plain white background communicates dimensions and finish. It does not communicate scale in a real room, how the piece sits with other furniture, or what it looks like in the light conditions a buyer will actually experience it in. Lifestyle imagery closes that gap. Buyers who can picture a piece in a room that looks like their own convert at higher rates than buyers who cannot.

The problem is that lifestyle shoots are expensive. Booking a studio space, sourcing complementary furniture and props, hiring a photographer, and retouching the results costs significantly more than a standard product shoot. For brands with large catalogues, doing this for every piece is not economically viable.

How do furniture brands create room scene imagery without reshooting?

The approach this brand uses starts with the existing white-background studio shot. That image goes into a cnaps studio pipeline alongside three text inputs describing the target room environments. The pipeline segments the furniture piece automatically using SAM2, composites it into each room setting, and produces three distinct lifestyle images in one run.

The furniture piece is extracted cleanly from its original background. Each room scene is generated around it with consistent lighting and spatial composition. The three outputs are reviewed in Image Compare before download.

How the cnaps studio pipeline works

This pipeline runs 20 nodes across a segmentation stage and three scene generation branches.

Step 01: Image Loader (Input)
The studio product shot loads into the pipeline. JPEG or PNG.

Step 02: Text Input x3 (Input)
Three text inputs describe the target room environments: for example, "Scandinavian living room, natural light, oak floor" and "dark bedroom, warm lamp light, linen bedding" and "home study, mid-century modern, bookshelves." Each text input drives one scene generation branch.

Step 03: Segmentation using SAM2 (AI Model)
SAM2 segments the furniture piece from the studio background automatically, producing a clean mask of the object. This mask is used in the compositing steps downstream. SAM2 handles complex furniture silhouettes including legs, arms, and open-frame structures accurately.

Step 04: Conditional Color Pick by Class (Tool)
The segmentation output is processed by class-based color selection to refine the furniture mask and isolate the exact pixels that belong to the piece.

Step 05: Gaussian Blur (Tool)
A gaussian blur is applied to the mask edges to produce a natural shadow boundary. This prevents the hard-edge halo effect that appears when a cleanly masked object is composited into a scene without edge softening.

Step 06: Image Multiply (Tool)
The blurred mask is composited with the background layer using multiply blending, producing a natural shadow under the furniture piece in each room scene.

Step 07: Image Add (Tool)
The furniture element and the shadow layer are combined using additive blending to produce the composited base image for each scene.

Step 08: QWEN Image Edit x3 (AI Model)
Three separate QWEN Image Edit passes run in parallel, one per room environment. Each pass receives the composited base image and its corresponding room description text input and generates the full room scene around the furniture piece. Processing time is approximately 1 to 3 minutes per pass.

Step 09: Color Inverse (Tool)
A color inverse step processes an intermediate output for one of the scene branches, used for specific lighting and tone adjustments in that variant.

Step 10: Image Viewer x5 and Image Compare (Output)
The three room scene outputs, plus intermediate outputs from the segmentation and compositing steps, are displayed for review. Image Compare shows the original studio shot alongside each room variant before download.

What makes the shadow and compositing steps important?

A furniture piece dropped into a room scene without shadow treatment looks pasted in. The Gaussian Blur and Image Multiply steps produce a soft, directional shadow under the piece that matches the lighting of the generated room environment. This is what makes the composited result look like the furniture belongs in the room rather than sitting on top of it.

Which furniture types work best with this pipeline?

The pipeline works best for freestanding furniture with a defined silhouette: sofas, chairs, tables, shelving units, and beds. SAM2 handles open-frame structures and complex leg geometry accurately.

Built-in furniture, flat-pack items sold unassembled, and very large pieces that occupy most of the studio frame are less suited to this approach. For these, the compositing result depends heavily on how well the original studio shot isolates the piece.

Try this pipeline on cnaps.ai

Cnaps.ai is a no-code visual platform for building and running multi-model AI pipelines. Fork the furniture room scene pipeline from the community and run it against your own product catalogue. No code required.

View and fork this pipeline on cnaps.ai

Frequently asked questions

Can I generate more than three room settings from the same product photo?

The current pipeline configuration produces three room variants per run. To generate additional settings, run the pipeline again with different room description text inputs. The same studio product shot is used as input each time.

Does the furniture piece change when it is placed in a room scene?

No. SAM2 segments the furniture from the original studio shot and the piece is composited into each room setting without alteration. The dimensions, finish, and colour stay accurate across all three variants. QWEN Image Edit generates the room environment around the furniture, not the other way around.

How does the pipeline handle furniture with complex shapes like open shelving or chairs with gaps?

SAM2 is designed for complex object segmentation and handles open-frame structures, chair legs, and shelving units accurately. The Conditional Color Pick by Class step refines the mask after segmentation to ensure that gaps in the furniture silhouette are handled correctly in the compositing stage.

What room environments can the pipeline generate?

The room environment is defined by the text input. Any room style that can be described in text can be generated: contemporary, traditional, Scandinavian, industrial, maximalist, and so on. The text input can specify lighting conditions, flooring, complementary furniture, and other environmental details that QWEN Image Edit incorporates into the scene.