The motor disability in Cerebral palsy (CP) is multifaceted and includes spasticity, weakness, and movement dysfunction, which is secondary to muscular contraction and skeletal abnormalities. The motor disorders of such individuals are gait disorders that are primarily related to defects in posture and movements that lead to activity restrictions including the ability to walk. Today, clinical gait analysis is commonly used to assess and diagnose the specific difficulties in the mobility of patients with CP. Though these systems provide better spatiotemporal and kinematic analysis of motion their high costs for implementation and immobility allow their usage only in lab settings. Also marker-based systems need markers that can negatively affect the natural gait patterns of the subject. Because of these drawbacks, the clinical analysis is inadequate for CP patients rehabilitation. New markerless motion capture technologies show great potential with advancements in modern techniques. Consequently, this study intends to use markerless deep learning models for gait analysis that will make useful findings for individual rehabilitation plans and clinical practice decisions.