<p>Flexible sensors with a porous hydrogel structure have attracted enormous attention for their extensive potential prospects in the fields of wearable electronics and human–machine interaction (HMI). Nevertheless, these sensors encounter significant challenges in synchronously achieving breathability, favorable biocompatibility, and high strain/pressure sensitivity. Herein, various sodium alginate (SA)-based hydrogel sponges are prepared by a facile one-pot and directional freezing process for comparative studies. The obvious oriented and uniform pore structures can provide a spatial anisotropy dispersion framework for the conductive fillers, and the effective connectivity among these interface structures facilitates the rapid deformation and recovery of the entire structure. Compared to other hydrogel sponges, the SA/silk fibroin (SF) hydrogel sponge demonstrates a more consistent pore distribution, preferable size uniformity, mechanical strength, deformation capability, and porosity properties. The designed SA/SF-based strain sensor with a satisfied gauge factor of 1.69 and durability can precisely monitor the physiological signals produced by humans. It exhibits favorable pressure sensitivity (− 1.10&#xa0;kPa⁻¹) and cyclic stability (more than 3750 cycles). Importantly, the strain and pressure stimuli can be discriminated by the proposed sensor via generating opposite resistance signals. Especially, combined with a microprocessor and Morse code, the prepared sensor can monitor different stimuli to translate into regular positive or negative signals, thus precisely performing effective information encoding and translation with the assistance of a deep learning model (higher accuracy of 99.2%). Consequently, our study proposes an effective strategy for fabricating strain and pressure sensors with good comprehensive properties in the fields of personalized healthcare and HMI.</p> Graphical Abstract <p></p>