The automatic recognition of food on images has numerous interesting applications, including nutritional tracking in medical cohorts. The problem has received significant research attention, but an ongoing public benchmark to develop open and reproducible algorithms has been missing. Here, we report on the setup of such a benchmark using publicly available food images sourced through the mobile MyFoodRepo app. Through four rounds, the benchmark released the MyFoodRepo-273 dataset constituting 24,119 images and a total of 39,325 segmented polygons categorized in 273 different classes. Models were evaluated on private tests sets from the same platform with 5,000 images and 7,865 annotations in the final round. Top-performing models on the 273 food categories reached a mean average precision of 0.568 (round 4) and a mean average recall of 0.885 (round 3). We present experimental validation of round 4 results, and discuss implications of the benchmark setup designed to increase the size and diversity of the dataset for future rounds.

First, a meshless simulation method is presented for multiphase fluid-particle flows with a two-way coupled Smoothed Particle Hydrodynamics (SPH) for the fluid and the Discrete Element Method (DEM) for the solid phase. The unresolved fluid model, based on the locally averaged Navier Stokes equations, is expected to be considerably faster than fully resolved models. Furthermore, in contrast to similar mesh-based Discrete Particle Methods (DPMs), our purely particle-based method enjoys the flexibility that comes from the lack of a prescribed mesh. It is suitable for problems such as free surface flow or flow around complex, moving and/or intermeshed geometries and is applicable to both dilute and dense particle flows. Second, a comprehensive validation procedure for fluid-particle simulations is presented and applied here to the SPH-DEM method, using simulations of single and multiple particle sedimentation in a 3D fluid column and comparison with analytical models. Millimetre-sized particles are used along with three different test fluids: air, water and a water-glycerol solution. The velocity evolution for a single particle compares well (less than 1% error) with the analytical solution as long as the fluid resolution is coarser than two times the particle diameter. Two more complex multiple particle sedimentation problems (sedimentation of a homogeneous porous block and an inhomogeneous Rayleigh Taylor instability) are also reproduced well for porosities 0.6 <= \epsilon <= 1.0, although care should be taken in the presence of high porosity gradients. Overall the SPH-DEM method successfully reproduces quantitatively the expected behaviour in the test cases, and promises to be a flexible and accurate tool for other, realistic fluid-particle system simulations.