receiving one or more images, wherein the one or more images includes a plurality of frames and no more than a partial annotation of the one or more images;

determining a loss component associated with a image segmentation model;

for each pixel of the one or more images, identifying a majority class and identify a cross-entropy loss between a network output and a target;

randomly selecting pixels associated with the one image and select a second set of pixels to compute a super pixel loss for each pair of pixels;

summing corresponding loss associated with each pair of pixels;

for each corresponding frame of the plurality of frames of the image, computing a positive flow loss, a negative flow loss, a contrastive optical flow loss, and a equivariant optical flow loss;

computing a final loss including a weighted average of the positive flow loss, the negative flow loss, the contrastive optical flow loss, the equivariant optical flow loss, the cross entropy loss, the super pixel loss, and foreground loss;

updating a network parameter associated with the image segmentation model in response to the final loss; and

outputting a trained neural network utilizing the updated network parameter in response to exceeding a convergence threshold.