PROBLEM TO BE SOLVED: To precisely grasp how much a user concentrates.SOLUTION: A concentration degree estimation device is provided with: an imaging part (26) which images a user (U1, U2, or U3)； a detection part (13) which detects indoor environment information associated with an environment in a room (r1) where the user is present； an estimation part (66) which uses an estimation model (67) based upon the image of the user that the imaging part (26) images and the indoor environment information that the detection part (13) detects, so as to estimate how much the user (U1, U2, or U3) concentrates； a control part (14) which controls, based upon the estimation result of the estimation part (66), first air-conditioning operation of an air-conditioning device (10) targeting the inside of the room (r1) as an object to be air-conditioned so that the user (U1, U2, or U3) concentrates more； and a model update part (68) which grasps how much the user (U1, U2, or U3) actually concentrates from an image that the imaging part (26) takes after the first air-conditioning operation based upon the estimation result of the estimation part (66) is performed, and then learns an estimation model (67) based upon the grasping result.SELECTED DRAWING: Figure 4COPYRIGHT: (C)2020,JPO&INPIT【課題】ユーザの集中度合いを精度良く把握する。【解決手段】ユーザ（U1,U2,U3）を撮像する撮像部（26）と、ユーザが居る室内（r1）の環境に関する室内環境情報を検知する検知部（13）と、撮像部（26）が撮像したユーザの画像と検知部（13）が検知した室内環境情報とに基づいて、推定モデル（67）を用いてユーザ（U1,U2,U3）の集中度合いを推定する推定部（66）と、推定部（66）の推定結果に基づいて、ユーザ（U1,U2,U3）の集中度合いが上昇するように、室内（r1）を空調対象とする空気調和装置（10）の第１空調動作を制御する制御部（14）と、推定部（66）の推定結果に基づく第１空調動作が行われた後の撮像部（26）が撮像した画像から各ユーザ（U1,U2,U3）の実際の集中度合いを把握し、その把握結果に基づいて、推定モデル（67）を学習させるモデル更新部（68）とを設ける。【選択図】図４
