Detecting diffuse radio emission, such as from haloes, in galaxy clusters is crucial for understanding large-scale structure formation in the universe. Traditional methods, which rely on X-ray and Sunyaev-Zeldovich (SZ) cluster pre-selection, introduce biases that limit our understanding of the full population of diffuse radio sources. In this work, we provide a possible resolution for this astrophysical tension by developing a machine learning (ML) framework capable of diffuse emission detection with only radio observations, using a limited real data set like those from the Murchison Widefield Array (MWA). We generate for the first time radio halo images using Wasserstein Generative Adversarial Networks (WGANs) and Denoising Diffusion Probabilistic Models (DDPMs), and apply them to train a neural network classifier independent of pre-selection methods. The halo images generated by DDPMs are of higher quality than those produced by WGANs. The diffusion-supported classifier with a multihead attention block achieved the best average validation accuracy of 95.93 per cent over 10 runs, using 36 clusters for training and 10 for testing, without further hyperparameter tuning. Using our classifier, we rediscovered 9/12 haloes (75 per cent detection rate) from the MeerKAT Galaxy Cluster Legacy Survey Catalogue, and 5/8 haloes (63 per cent detection rate) from the Planck SZ Catalogue 2 within the GaLactic and Extragalactic All-sky MWA (GLEAM) survey. In addition, we identify 11 potential new haloes, minihaloes, or candidates in the COSMOS field using Chandra-detected clusters in GLEAM data. This work demonstrates the potential of ML for less-biased detection of diffuse emission and provides labelled data sets for further study.