nestedxml fetched converted to nested dict. would like advice how to do it efficiently

Question

I'm parsing nested xml from fetched from postgres and packing it into nested dict. In the process I'm creating multiple temp variable to pass value after condition is met. I'm also using nested loop. There is a lot of rows that needs to be fetched from postgres. This is the solution that I came up with but would like to know if this could done is efficient way without complicating. I can't copy paste the code so I rewrote this on here but this seems to work even if there's typing mistakes. Looking for advice to better this code.

innerscore, innerpop = {},{}
    stmt = "....."
    with conn.cursor(name="segcurs", cursor_factory=psycopg2.extras.RealDictCursor,scrollable=False, withhold=False) as segcurs:
        segcurs.intersize=1000
        segcurs.execute(stmt)
        for row in segcurs:
        tree = ET.formstring(row['per'])
        for child in tree:
            if child.tag ="score":
                temp_d = innerscore
            if child.tag = "pop":
               temp_d = innerpop
            for child1 in child:
               name = child1.tag
               value = child1.text
            if not name in temp_d:
                temp_d[name] = []
                temp_d[name].append(value)
            else:
                temp_d[name].append(value)
    
    perf={}
    perfs['score'] = innerscore
    perfs['population'] = innerpop
    row = <"per"><"score"><"s1">1</"s1"><"s2>2</"s2"></"score">,<"pop"><"p1">1</"p1"><"p2>2</"p2"></"pop"></"per"> (long xml)
    perfs={"score":{"s1":[],"s2":[]}, "population":{"pop1":[], "pop2":[]}}