Constitutive models are used to describe the mechanical behavior of materials (Karev et al., 2020;Nadai, 1963;Puzrin, 2012). These consist of mathematical equations that can be used to relate physical quantities and are defined by material-specific constants (Davis & Selvadurai, 2005). The design of civil engineering structures relies on the precise calibration of these constants (Alonso et al., 2010;. Triaxial tests are commonly carried out to obtain the constitutive parameters of geomaterials (Jaeger et al., 2007). It is generally assumed that the triaxial test is a representative volume test, meaning that the developed stresses and strains are uniform in the sample. Accurate detection of a sample's strain is important for the determination of the mechanical properties. Linear variable differential transformers (LVDT) and strain gauges are common technologies to measure the strain response. The drawback is that the measurement is limited to the point or small zone where they are installed, forcing the assumption of a homogeneous sample response in accordance with the representative volume assumption. However, recent numerical and experimental studies have shown strain localization in the samples, even at early test stages (McBeck et al., 2019;Van der Baan & Chorney, 2019). This implies that when point measurement methods are used, the mechanical parameters obtained are dependent on the sensor location.