Skip to main content

Table 3 Failure cases: sample input and output given by the framework

From: Semantic context driven language descriptions of videos using deep neural network

Figure No.

Ground-truth

Model_1

Model_2

Model_3

Figure 6a

{A car running from the police},{A guy is riding too fast in his bike.},{A man is driving backward and spins the car around.}

{A man is playing a guitar}

{A car is going up }.

{A car is chasing a car}.

Figure 6b

{A dog climbed into a clothes washing machine.}, {A bull dog is jumping into a washing machine.}, {The puppy went into the dryer.}, {The dog crawled into the dryer.}

{A man is putting some vegetables in a pan}

{A man is beating a concrete into a water}

{A man is making a fancy dish}.

Figure 6c

{ Airoplane in the Air}, {The plane took off from the runway.}, {An airplane is taking off.},{the person going on the airplane}

{A man is riding a bike}.

{A woman is pushing a rock}.

{A woman is running in the air }