FVLBI

Cataloguing a wardrobe

A garment is photographed flat, in even light, from above. FVLBI reads its colour and identifies the material, covering cotton, wool, silk, linen, cashmere, leather, suede, denim, polyester, nylon and velvet. It infers formality and season from the cut, and reads patterns where they are present, from solids through florals, stripes, plaid and geometrics.

Every reading can be corrected by hand. Pieces carry a state, so the wardrobe knows what is clean and what is in the wash, and they can be removed when they are sold or given away. The record stays accurate to what someone actually owns.

Colour

FVLBI places each person in one of twelve seasonal palettes: Bright, True and Light Spring; Light, True and Soft Summer; Soft, True and Deep Autumn; Deep, True and Bright Winter. Each palette is a coherent set of colours that suits a particular combination of underlying tones.

The placement is read from a short selfie and a body photograph, taking skin undertone, hair colour and eye colour into account. If the lighting is too uneven for a confident reading, the app asks for a retake. Colouring shifts slowly, so the analysis can be run again from settings after a significant change such as a new hair colour.

The palette guides rather than rules. It carries roughly fifteen per cent of the reasoning behind a composition, and pieces outside it stay in rotation. In-palette pieces are favoured where they fit, not enforced.

Composing an outfit

Outfits are assembled only from pieces already in the wardrobe. A composition accounts for the weather forecast, the formality of the day ahead, and what is currently in the wash. The day’s suggestion is held stable rather than changing on every visit, and it can be biased toward a particular colour or style.

Weather

A composition is shaped by the day’s forecast, read from local temperature and conditions at the start of the day. Every catalogued piece carries a season tag, and the temperature decides what is surfaced and what is held back: wools, knits and outerwear on cold days, cottons and linens on warm ones.

Forecasts are wrong sometimes. A composition holds to the reading taken when it was made rather than shifting underneath the wearer, and a fresh one is a tap away if the day turns.

Previewing the result

Once an outfit is composed, FVLBI renders a figure matching the wearer’s body shape and proportions, wearing the exact pieces from their wardrobe. The output is photographic rather than illustrative. Standard previews are stills at 1024 pixels on the long edge; higher tiers add 2048-pixel stills and short video previews.

Solid colours, classic silhouettes and clean construction render most accurately. Heavily patterned pieces, very technical fabrics and complex tailoring render less precisely.

Platforms and availability

FVLBI runs in the browser today at fvlbi.com, on desktop and on mobile, with a native iOS application in preparation. Wardrobes and compositions are held against an account rather than a device, so the same wardrobe is there wherever it is opened.

Standard previews are available on every plan. Higher tiers add larger renders and short video previews, for the cases where a still image does not settle the question.

The product is operated by Jupetto (Pty) Ltd, which handles its engineering, infrastructure and support.