Integrals refining convex inequalities (Q1988590)

The authors prove that, if \(\Phi:\mathcal{B}(\mathcal{H})\rightarrow\mathcal{B}(\mathcal{K})\) is a normalized positive linear map, \(A\in\mathcal{B}(\mathcal{H})\) is a self-adjoint operator with the spectrum in \(J\), and \(f:J\rightarrow\mathbb{R}\) is a convex function, then \begin{align*} f\left(\frac{\langle\Phi(A)x, x\rangle+\langle\phi(A)y, y\rangle}{2}\right) &\le \int_0^1f\big(t\langle\Phi(A)x, x\rangle+(1-t)\langle\Phi(A)y, y\rangle\big)\,\mathrm{d}t\\ &\le \left\langle\Phi\left(\int_0^1f\big(tA+(1-t)\langle\Phi(A)y, y\rangle\big)\,\mathrm{d}t\right)x, x\right\rangle\\ &\le \frac{\langle\Phi(f(A))x, x\rangle+\langle\Phi(f(A))y, y\rangle}{2} \end{align*} for any \(x, y\in \mathcal{H}\) with \(\|x\|=\|y\|=1\). Also, the authors state several consequences of this theorem to scalar functions, including a refinement of the discrete version of Jensen's inequality and some versions of the Hermite-Hadamard inequality. Finally, the authors present some nice graphical explanations.