複数人の生成

AI画像生成の場合、基本的に１人の画像が多いのですが、複数人の登場が出来ないのかと言うと決してそんな事はありません。

まず基本画像。
1人の女性がスタバのコーヒーを持って散策中。

プロンプトはこんな感じ。
A tree-lined street in a business district in Tokyo.

This photo is a close-up of a Japanese actress (30-15 years old) walking down an avenue lined with yellow broad-leaved trees.

She is drinking from a Starbucks cup in one hand. She is wearing a white knit sweater, a beige knee-length A-line skirt, black pantyhose, and brown leather shoes. Her hair is short, light brown and wavy.

The setting sun is backlighting her brightly. Ginkgo leaves are scattered on the cobblestone road

これを基本として人数を増やしたプロンプトを書きました。
解りやすい様に、それぞれの衣装やポージングに明らかな違いを出しました。

プロンプト

A pedestrian-only tree-lined street in Tokyo's business district.

This photo shows three Japanese actresses (30-15 years old) chatting while walking along a street lined with yellow broad-leaved trees.

The woman on the left is drinking a Starbucks short cup in one hand. She is wearing a white knit sweater, a beige A-line knee-length skirt, black pantyhose, and brown leather shoes. She has light brown wavy short hair.

The woman in the center with black bob hair is smiling with her hand over her mouth. She is wearing a long-sleeved mini dress with a beige base and a crimson, green, and black check pattern, and black short boots.

The woman on the right has her hair in a ponytail. She is wearing a green jacket over a white blouse, gray culottes, and green basketball shoes. She is talking with the woman in the center with her arms outstretched, and seems to be having fun.

A bright sunset is illuminating them.

Many ginkgo leaves are scattered on the cobblestone road

一応3人で設定していますがアスペクト比を横長にすると、余ったスペースにもう一人ねじ込んでくる場合もあります。そこら辺はガチャですね…

まあ、そんな感じで人数を増やせるのですが、団体の場合は服装指定も統一プロンプトで済みますから更に楽かもしれません。

皆様も、是非上記のプロンプトを改変しながらお楽しみください。